THE STORY OF 
 
 ARCHITECTURE 
 
 IN 
 
 OXFORD STONE 
 
 BY 
 
 E A. GREENING LAMBORN 
 
 OXFORD 
 
 AT THE CLARENDON PRESS 
 1912 
 
HENRY FROWDE 
 
 PUBLISHER TO THE UNIVERSITY OF OXFORD 
 
 LONDON, EDINBURGH, NEW YORK 
 
 TORONTO AND MELBOURNE 
 
PREFACE 
 
 
 
 THIS book is not meant to be an addition to the already 
 numerous guides to the individual buildings of Oxford. 
 As such it could find little justification, since every 
 important building has at some time or other been the 
 subject of a book, in which its history and its architectural 
 features have been exhaustively treated, and there are 
 besides several guides to the city as a whole, in which 
 accounts are given of its chief architectural details with 
 the dates and styles of all the colleges and churches. 
 
 Not Oxford's buildings, but the science of architecture 
 illustrated by them, is the subject of this essay. 
 
 As a rule, writers on English architecture draw their 
 examples from buildings scattered broadcast over England ; 
 the majority of students must, therefore, be content to 
 make acquaintance with their details through the medium 
 of photographic illustrations, drawings, and descriptions, 
 which are at best a poor substitute for the real thing. 
 Now Oxford, a unique city in so many respects, is unique 
 in this, that all the great architectural types are represented 
 in her buildings. It is true that our examples of Classic 
 architecture are but poor imitations of the stately 
 porticoes of Greece and Rome, but they will still serve 
 to illustrate the mechanical principles and the ornamental 
 details of the ancient building systems ; of every stage 
 of mediaeval architecture Oxford possesses examples as 
 representative of the best work as are to be found any- 
 where in England ; the buildings of the great Renaissance 
 
 A2 
 
 274121 
 
4 PREFACE 
 
 architects are not better represented in London itself 
 than in the streets of our own city ; and finally, it was 
 Oxford that saw Loth the last efforts of expiring Gothic 
 and the first attempts at the revival of the mediaeval 
 style. 
 
 Here, then, is an opportunity to approach the study of 
 architecture with buildings of every period at hand for 
 illustrations ; I have tried to show how they may be 
 used to illustrate the development of the science from 
 primitive to modern times. 
 
 My main purpose has been less to describe the charac- 
 teristics of the work of different dates than to attempt 
 to trace through the successive styles a continuous line 
 of evolution. Therefore, minute descriptions of details 
 that the reader may observe for himself are unnecessary ; 
 the aim was rather to inquire into their origins and 
 functions, and to follow the history of their development. 
 
 Ability to .recognize the work of different periods in 
 an old building, and to trace in chronological sequence 
 the history of the structure is not very difficult to acquire, 
 and adds greatly to one's capacity for feeling the mysteri- 
 ous charm of ancient things. But it involves no more 
 real knowledge of architecture than the recognition of 
 the signs of age in a human being involves a knowledge 
 of physiology. The deeper intellectual interest of the 
 science is to be found, not in the naming and classification 
 of details, but in the study of their functions, and of the 
 adaptations of their forms to the ends they are intended 
 to serve ; it is a physiological interest. 
 
 This book is, in its small way, an attempt to apply to 
 the study of architecture the methods of modern biology : 
 to trace the origin and development of architectural 
 forms ; to explain their functions and their inter- 
 
PREFACE 5 
 
 relations ; and to seek out the causes that modified them 
 and controlled their development. It proceeds on the 
 premisses that architecture is analogous to an organic 
 growth, that its study should therefore be approached 
 from a genetic and evolutionary standpoint, the student 
 seeking to explain its development by reference to the 
 changes in human circumstances, just as the biologist 
 seeks to explain the development of species from simple 
 to complex, not by the theory of special creations, but 
 by the effects of environment upon the organism. The 
 old botanist was content when, from observation of out- 
 ward resemblances, he had referred a plant to its natural 
 order ; to the post-Darwinian, classification is not the 
 association of like forms, but the relating of species to 
 a common ancestor ; he is not satisfied until he has 
 explained the differences between related species by 
 reference to the varying circumstances of their environ- 
 ments. The belief that each species sprang into existence 
 by a separate act of the Creative Mind has given place 
 to the nobler theory of evolution from a single primitive 
 form of life. 
 
 I have tried to apply the evolutionary method to the 
 study of architecture, and to show that in the history 
 of building, as in that of organic life, there is a single 
 primitive type from which all later forms were evolved ; 
 that all the varied styles belong to one or other of a few 
 great branches ; that the line of progress is from simple 
 to complex, from the lowly organism to the high ; from 
 the undifferentiated form to the specialized, from the 
 rudimentary to the highly developed ; and that the 
 changes that mark that progress were the results, not of 
 changing fashions or of the caprice of individual architects, 
 but of the pressure of new circumstances. 
 
6 PREFACE 
 
 Of the many writers to whom I am more or less in- 
 debted, I owe the largest amount of gratitude to three : 
 Mr. Garbett, whose Principles of Design in Architecture, 
 probably the most philosophically written treatise on 
 the subject, first suggested to me the idea of evolution 
 in architecture ; Mr. Bond, who, in his great work, 
 Gothic Architecture in England, has analysed that system 
 with a completeness that must make all later writers his 
 debtors ; and Mr. Jackson, whose book, Reason in 
 Architecture, has been full of suggestion for me, and whose 
 accounts of S. Mary's and Wadham have provided me 
 with much information. I have to thank the Dean of 
 Christ Church and the Wardens of Merton and New 
 Colleges for permission to take photographs. Finally, 
 I owe to Mr. R. W. Chapman, and to my friend 
 Mr. C. R. L. Fletcher, most grateful acknowledgement 
 of wise and helpful criticism in the manuscript stage 
 of the book. 
 
CONTENTS 
 
 PAGE 
 
 INTRODUCTION . . . . 13-17 
 
 I. WHAT is ARCHITECTURE ? . . .13 
 
 II. ROOF-MAKING .... .14 
 
 PART I 
 
 THE HISTORY OF ARCHITECTURE 19-92 
 
 CHAP. 
 
 I. ANCIENT ARCHITECTURE .... 19 
 
 II. THE INVENTION OF THE ARCH ... 33 
 
 III. ROMANESQUE ARCHITECTURE ... 42 
 IV. GOTHIC ARCHITECTURE . . . .55 
 
 V. THE RENAISSANCE AND AFTER ... 72 
 VI. THE GOTHIC REVIVAL .... 80 
 
 PART II 
 
 THE GRAMMAR OF ARCHITECTURE 93-285 
 INTRODUCTORY ...... 93 
 
 I. THE PLAN AND ELEVATION ... 94 
 II. THE VAULT .109 
 
 III. THE BUTTRESS 124 
 
 IV. THE ARCH . 131 
 
8 CONTENTS 
 
 CHAP. PAGE 
 
 V. THE PIER ...... 143 
 
 VI. THE ROOF ...... 159 
 
 VII. THE WALL 174 
 
 VIII. THE DOORWAY 186 
 
 IX. THE WINDOW 195 
 
 X. THE WINDOW (continued) . . .214 
 
 XI. THE WINDOW (concluded) . . . 224 
 
 XII. ARCHITECTURAL ORNAMENT . . . 233 
 
 XIII. ARMORY IN ARCHITECTURE . . 256 
 
 XIV. THE INTERIOR ARRANGEMENTS IN MEDIAEVAL 
 
 BUILDINGS ..... 266 
 
 INDEX 286 
 
LIST OF ILLUSTRATIONS 
 
 FIG. PAGE 
 
 1. Doric Order ....... 25 
 
 2. Ionic Order ....... 27 
 
 3. Corinthian Order 29 
 
 4. Classic Mouldings . . . . . 3 1 
 
 5. Sheldonian Theatre 37 
 
 6. Barrel Vault ....... 38 
 
 7. Groined Cross-vault 39 
 
 8. Saxon Arch in Cathedral (c. 727) ... 45 
 
 9. The Castle Mound (c. 912) 47 
 
 10. S. Michael's Tower (c. 1071) .... 48 
 
 11. Iffley Church, West Front (c. 1 1 60) . . . 51 
 
 12. The Castle and Mill 53 
 
 13. Interior of Cathedral ...... 59 
 
 14. S. Giles's Church (c. 1220) . . . . . 61 
 
 15. The Chapel, Merton College (c. 1297) ... 63 
 
 1 6. The Cloisters, New College (c. 1386) ... 64 
 
 17. The Divinity School (c. 1450) .... 67 
 
 1 8. Vault of Divinity School 69 
 
 19. The Quadrangle, Oriel College (1637) ... 71 
 
 20. The Quadrangle, S. John's College (c. 1 630) . . 73 
 
 21. Clarendon, Sheldonian, and Old Ashmolean Buildings 75 
 
 22. All Saints' Church (c. 1709) .... 77 
 
 23. Seventeenth-century House, Balliol College . . 81 
 
 24. Bishop King's Palace ...... 83 
 
 25. Exeter College Chapel 85 
 
 26. New Examination Schools 87 
 
 27. Central Spire of Cathedral (c. 1250) . . .96 
 
 28. The Monastic Plan 101 
 
io LIST OF ILLUSTRATIONS 
 
 FIG. PAGE 
 
 29. Typical Gothic Elevation . . . . .107 
 
 30. Crypt of S. Peter's in the East (c. 1120) . . in 
 
 31. Vault of Chancel, S. Peter's . . . .112 
 
 32. Vault-rib : Primitive Type . . . .114 
 
 33. Vault-rib: Typical Norman . . . .114 
 
 34. Vault-rib: Transitional . . . . 115 
 
 35. Vault-rib: Early English . . . . 115 
 
 36. Vault-ribs in Cathedral 117 
 
 37. Latin Chapel, Christ Church (c. 1350) . . . 119 
 
 38. Staircase, Christ Church (1640) . . . .121 
 
 39. Norman Buttress . . . . . .125 
 
 40. Iffley Church, East End (c. 1 1 60) . . . 127 
 
 41. Buttress, S. Mary Magdalene (c. 1337) . . 128 
 
 42. Buttress, Divinity School (c. 1450) . . .130 
 
 43. Lancet Arch 134 
 
 44. Equilateral Arch . . . . . -134 
 
 45. Drop Arch . . . . . .134 
 
 46. Ogee Arch 1 34 
 
 47. Four-centred Arch . . . . . 135 
 
 48. Late Four-centred Arch . . . . 135 
 
 49. Fourteenth-century Doorway, Witney . . 135 
 
 50. Interior of S. Mary's (c. 1488) . . . -137 
 
 51. Priest's Door, S. Thomas's (c. 1220) . . -139 
 
 52. Arch Construction, Roman Method . . .141 
 
 53. Arch Construction, Mediaeval Method . . 142 
 
 54. Arch and Piers, S. Mary's, A. D. 1488 . . . 144 
 
 55. Thirteenth-century Piers, S. Peter's . . . 146 
 
 56. Thirteenth-century Pier, Holywell Church . . 147 
 
 57. Tower Piers, Merton College (c. 1290) . . 149 
 
 58. Cushion Capital, Cassington (c. 1140) . -150 
 
 59. Scallop Capital (c. 1160) 150 
 
 60. Capital in Cathedral (c. 1 1 80) . . . .151 
 
LIST OF ILLUSTRATIONS n 
 
 FIG. PAGE 
 
 61. Pier of Chancel Arch, Elsfield (c. 1200) . 153 
 
 62. Capital (c. 1220) 154 
 
 63. Capital, Merton College Chapel (c. 1280) . . 155 
 
 64. Pier, S. Mary's, A. D. 1488 . . ' . .156 
 
 65. Fifteenth-century Base, S. Mary's, A.D. 1488 . 158 
 
 66. S. Peter's in the East 162 
 
 67. The City Wall 167 
 
 68. S. Mary Magdalene Church (c. 1337) . . .169 
 
 69. North Transept, Merton College Chapel (c. 1420) . 178 
 
 70. Norman String-course (c. noo) .... 179 
 
 71. Transitional String-course (c. 1180) . . . 179 
 
 72. Early English String-course (c. 1220) . .180 
 .73. Decorated String-course (c. 1300) . . . 180 
 
 74. Perpendicular String-course (c. 1450) . . .181 
 
 75. Linen Panelling (c. 1500) ..... 185 
 
 76. South Door, Iffley Church (c. 1150) . . .187 
 
 77. Thirteenth-century Doorway, Milton (c. 1240) . 189 
 
 78. Early Fourteenth-century Doorway, Dorchester . 190 
 
 79. Tower of the Five Orders . . . . .191 
 
 80. Fifteenth-century Doorway, Merton College Chapel, 
 
 A.D. 1424 ....... 192 
 
 81. Queen's College Gateway (c. 1710) . . .194 
 
 82. Norman Window, Sandford (c. noo) . . . 200 
 
 83. Late Norman Window, Cathedral (c. 1 1 86) . .201 
 
 84. Late Norman Window, Sutton (c. 1170) . . 202 
 
 85. Lancet Window (c. 1220) 203 
 
 86. The Germ of Tracery, S. Giles's (c. 1200) . . 206 
 
 87. Early Tracery, Woodstock (c. 1240) . . . 207 
 
 88. Early Tracery, S. Giles's (c. 1265) . . .208 
 
 89. Geometrical Tracery, Merton College Chapel (c. 1297) 210 
 
 90. Very Early Bar-tracery, Northmoor Church (c. 1270) 211 
 
 91. Window, S. Michael's (c. 1260) . . . .213 
 
12 LIST OF ILLUSTRATIONS 
 
 FIG. PAGE 
 
 92. Window, S. Michael's (c. 1320) . . . .216 
 
 93. Window, S. Mary Magdalene (c. 1337) 2I 7 
 
 94. Window, S. Peter's in the East (c. 1360) . . 221 
 
 95. Fifteenth-century Window, S. Mary's . . 222 
 
 96. Jacobean Windows, S. John's College . . 227 
 
 97. Cowley Church ...... 229 
 
 98. Ruined Shrine of S. Frideswide (c. 1289) . -235 
 
 99. Chevron Ornament ...... 242 
 
 100. Chevron Ornament ...... 242 
 
 101. Norman Pellets 243 
 
 102. Dog-tooth Ornament . . , . . 243 
 
 103. Beak-heads ....... 244 
 
 104. Beak-heads, S. Ebbe's Church .... 244 
 
 105. Thirteenth-century Hollow Moulding . . . 245 
 
 106. Ball-flower Ornament, Prior Sutton's Tomb 
 
 (c. 1300) ....... 246 
 
 107. Fillets ........ 247 
 
 108. Canopies of Sedilia, Merton College Chapel (c. 1300) 249 
 
 109. Ogee Moulding (c. 1350) ...... 250 
 
 no. Double Ogee, Fifteenth Century . . . 250 
 in. Fifteenth-century Hollow Moulding . . .251 
 
 112. Sedilia, S. Mary's (c. 1488) . . . -253 
 
 113. Schools Gateway 254 
 
 1 14. The Ordinaries of Armory .... 259 
 
 115. Thirteenth-century Fireplace, Abingdon Abbey . 269 
 
 1 1 6. Chimney Turret, Abingdon Abbey (c. 1250) . 271 
 
 117. Wall-painting, South Leigh . . 277 
 
 118. Squints at Haseley ... . 279 
 
 119. Seventeenth-century Woodwork, Cumnor . .281 
 
 120. Flat Gravestone, Great Milton, Oxon. (c. 1250) . 283 
 
 121. Tomb, Waterperry, Oxen. (c. 1400) . . 284 
 
INTRODUCTION 
 
 I. WHAT is ARCHITECTURE ? 
 
 ' PIETY, Socrates, everybody knows what that means.' 
 So, probably, everybody knows what architecture means 
 until he conies to define it. Then most people would 
 find, that while they were able at once to recognize 
 whether or not a building possessed the character of 
 architecture, they could much less readily declare in 
 what attributes that character consisted. One could 
 easily make a list of buildings certainly architectural, 
 and another of buildings certainly not so, but one might 
 have considerable difficulty in stating the grounds upon 
 which a particular building was placed in either list. 
 It will be neither uninteresting nor unprofitable to pause 
 at the outset and try to find out definitely what element 
 it is in certain buildings that invests them with the 
 quality of architecture. 
 
 It has been said that, strictly speaking, all building 
 is architecture ; it might as truly be said that all writing 
 is literature. All building has for its prime purpose the 
 satisfaction of the first physical need of civilized man 
 the provision of shelter for himself and his belongings : 
 so utility is a characteristic common to all. But while 
 a mere builder is concerned wholly with the practical 
 uses his work is designed to serve, and is content according 
 to the degree in which his building answers its purpose, 
 the architect is not content to provide for physical needs 
 alone ; another ideal disputes with utility the possession 
 
H INTRODUCTION 
 
 of .his mind, and his building not only satisfies the physical 
 desire for comfort and convenience, but gratifies also 
 the higher human instinct by which man craves naturally 
 for seemliness and dignity in his surroundings. 
 
 In those buildings to which we apply the term architec- 
 tural there is the recognition of the deep truth that man 
 doth not live by bread alone, that a building is a part 
 of man's spiritual environment as well as a shelter for his 
 body, and that the dignity of humanity demands for 
 a human dwelling-place a certain excess of design and 
 workmanship beyond what is required to produce a con- 
 venient and comfortable building. That recognition 
 I believe to be the fundamental characteristic of what we 
 call architecture, the touchstone by which it may be 
 distinguished from mere building. Building is the art 
 of constructing animal dwellings ; it is not confined to 
 the human species. Architecture arose when man first 
 began to realize his higher nature ; it is the art of con- 
 structing buildings that satisfy human needs. 
 
 II. ROOF-MAKING 
 
 Architecture is the oldest of the arts, for we have seen 
 that it had its origin in the first need of civilized man, 
 a roof to shield him from the weather. To construct 
 a roof is the essential function of architecture. All the 
 various parts of a building, however complex, are built 
 in relation to the roof and subordinate to it ; the walls 
 or pillars that support it and enclose or subdivide the 
 space roofed by it, the buttresses that take its thrust, 
 the windows that light and the doorways that give access 
 to the space it covers, all are governed in their construction 
 by the dominating feature that fulfils the essential purpose 
 
INTRODUCTION 15 
 
 of the building. Fundamentally true is the figure of 
 speech that makes ' roof ' do service for a whole home- 
 stead. 
 
 This is no idle analysis, but the necessary elucidation 
 of a primary fact, upon the realization of which depends 
 the comprehension of the whole science of the architec- 
 tural styles. For the great building systems of the world 
 are distinguished, not by details, but by fundamentals 
 by their method of solving the first problem of architec- 
 ture, how to build a roof over a given space. 
 
 Three solutions have been evolved in the course of 
 human history ; and there are, therefore, only three great 
 architectural styles. 
 
 The first solution was arrived at before the dawn of 
 history ; it was the very simple and obvious plan of making 
 a horizontal roof of long poles covered with a layer of 
 rushes and supported upon stout posts. Even in the 
 twentieth century many a cart-shed or cattle-shelter in 
 a field-corner may.be seen roofed in this primitive fashion 
 with poles covered by brushwood. At first the water- 
 proof layer would be a solid mass of rushes with its sides 
 sloped to throw off the rain ; but soon would be dis- 
 covered the plan of constructing a sloping framework of 
 light poles covered with a skin of thatch to set upon the 
 flat roof, and so to save rushes. 
 
 Except that tiles and slates have been substituted for 
 thatch, this earliest type of roof, formed by horizontal 
 beams supporting a sloping framework of rafters, has per- 
 sisted until to-day, and covers most of our modern build- 
 ings. The horizontal principle is disguised by the sloping 
 tiles that hide the beams, but it is revealed in the flat 
 ceiling. Every building with a flat ceiling belongs to the 
 primitive or trabeated system ; it is constructionally 
 
16 INTRODUCTION 
 
 a ' post and lintel ' building, i. e. its roof is a dead weight 
 exerting a vertical pressure upon its supports. 
 
 The earliest architecture was, of course, timber con- 
 struction. But long before recorded history man learned 
 to substitute stone for, wood, and so to make more 
 dignified and durable dwellings. The change in material 
 was a great step forward ; it involved eventually a change 
 in constructive method ; but many ages had to pass 
 before this was realized and a new system invented. 
 Those who have seen Stonehenge or Weyland Smith's 
 cave will have noticed that in these primitive attempts 
 at stone architecture man followed the only plan of 
 roof-making he yet knew, merely substituting stone 
 columns for wooden posts, and great flat slabs for the 
 rush-covered beams. 
 
 But it was in Egypt, the cradle of civilization, that 
 architecture was developed into a science. Egypt had 
 little wood, but was rich in stone, and so her people were 
 forced into stone construction. The only method of 
 roof-making they knew was that of posts and beams ; 
 they therefore built their temples with flat stone roofs 
 supported by massive columns. They were not driven 
 by difficulties of material to evolve a new method, for 
 their stone was of such a nature that it could be extracted 
 in pieces of sufficient length ; great monoliths for the 
 columns, broad flags for the roof, and long masses for 
 lintels, were all available ; and labour was plentiful. 
 Young Greece found her architectural models in Egypt. 
 The spirit that was in her caused her to beautify them 
 and so to produce the Greek temples. But, again, no diffi- 
 culties of material forced her architects to devise a new 
 system. She perfected the style of lintel and column, 
 and she passed it on to Rome. 
 
INTRODUCTION 17 
 
 For a century or two the mistress of the world ran- 
 sacked her dominions for the monoliths and great stone 
 beams necessary for a trabeated system of architecture. 
 Then even her vast resources began to fail, and the builders 
 were driven to invent a system of roofing in which stones 
 of ordinary size could be used. They substituted the 
 arch and the vault for the lintel and the beam. So the 
 second great style of architecture came into existence, 
 the architecture of the round arch, the architectural 
 legacy of Rome to the modern world. 
 
 For a thousand years the round arch dominated 
 architecture. Then, just at the time when the modern 
 nations were coming into individual existence, the 
 builders of Western Europe discovered the pointed arch, 
 and applied it to a new system, the last and noblest, 
 Gothic architecture, which gave us 
 
 Minaret crowned S. Mary's and Magdalen tower and 
 Merton, 
 
 and all the dreaming spires that beautify our sweet city. 
 
 785 
 
PART I 
 THE HISTORY OF ARCHITECTURE 
 
 CHAPTER I 
 
 ANCIENT ARCHITECTURE 
 
 To understand the architectural styles represented in 
 Oxford's buildings some knowledge of ancient architec- 
 ture, and especially of Greek architecture, is necessary. 
 Rome learned the art of building from the Greeks, and 
 imparted it in turn to her successors with her own 
 amendments, upon which they again improved. From 
 Greek temple to Roman basilica, from basilica to early 
 Christian church, and from thence through rude Saxon 
 and barbaric Norman to the perfect loveliness of Ely 
 or Lichfield there is true organic unity. 
 
 We have in Oxford no examples of the handiwork of 
 the ancient civilizations, and no reproductions of it earlier 
 than the seventeenth century. Most architectural 
 handbooks, therefore, defer any account of the Classic 
 styles until they come to consider the work of the Renais- 
 sance architects. But the evolutionary method demands 
 that the mediaeval styles should be studied in the light 
 of those that preceded them. 
 
 I shall, therefore, devote the first section of the book 
 to a brief description of the ancient styles, to a discussion 
 of the difficulties inherent in the lintel system, and of the 
 manner in which these were overcome by the employment 
 of the arch, concluding with an account of the arcuated 
 
 B2. 
 
20 THE HISTORY OF ARCHITECTURE 
 
 system of the Roman builders, out of which the mediaeval 
 styles developed. 
 
 It will be convenient to interpolate here a few remarks 
 on the architectural vestiges of the Roman occupation 
 of this part of Britain. They are very few, and consist 
 only of foundations. It is highly improbable that a 
 Roman ever even set foot upon the actual site of Oxford. 
 Two great roads ran through the district from London 
 to Cirencester ; one, the Akeman Street, left Oxford 
 eight or nine miles to the south ; the other, passing 
 through Silchester, left it thirty miles to the north. 
 A small connecting road 1 ran north from Silchester, 
 through Dorchester to join the Akeman Street near 
 Bicester ; this road passed over Shotover Hill, and no 
 doubt many a Roman legionary, toiling up the steep 
 slope from Bayswater, must have halted on the crest to 
 breathe himself and to look westward down into the 
 misty hollow where now the city is cradled. In the 
 hedge near the brick works, just where the Roman road 
 crosses the old London highway, there is a great stone 
 that may well have been a Roman milestone. In the middle 
 of Otmoor there is another, now broken, marked on the 
 Ordnance Map as Joseph's stone. 
 
 Traces of Roman country-houses have been found at 
 many places along the lines of the three roads. These 
 sites are marked on the Ordnance Maps. The only places 
 at which remains are still to be seen are North Leigh, 
 where there is a fine Roman floor now protected by 
 a roof, and Widford, where the little church stands upon 
 
 1 It can still be followed, experto crede, from Bullingdon Green 
 to Dorchester ; but northwards it has been obliterated by the 
 
 S narries of Headington, or has sunk beneath the marshes of 
 tmoor. 
 
ANCIENT ARCHITECTURE 21 
 
 the site of a Roman villa, and has part of a tesselated pave- 
 ment in the floor of the chancel. 
 
 Excavations of the sites show that the villa was usually 
 built round three sides of a quadrangle, one block forming 
 the house, another the stores and stables, and the third 
 the kitchens, baths, and offices. The villas were inhabited, 
 not by Romans, but by Romanized Britons, and, in spite 
 of their tesselated floors and hypocausts, were rather of 
 the native than of the Roman type of architecture, being 
 built mainly of wood and of one story. 
 
 Judging by the remains preserved in our museums 
 there were very few buildings in Roman Britain at all 
 representative of Classic architecture. But the buildings 
 of the Renaissance period in Oxford will give a very good 
 idea of the principles and details of the ancient style, 
 and from them our examples will be drawn. 
 
 Greek architecture was a system of constructing 
 buildings with flat roofs supported by columns upon 
 which their weight exerted a vertical pressure. It is 
 therefore a post and lintel system, the posts and beams 
 of the older timber construction being replaced by 
 columns and long slabs of stone or marble. The 
 mechanics of the Greek temple are as simple as those 
 involved in the construction of a cart-shed ; if the reader 
 will imagine a billiard-table with its slate bed removed, 
 he will see in the remaining framework all the structural 
 parts of a Greek building. The legs of the table represent 
 the columns, the oblong frame supported by them 
 corresponds to the entablature upon which the flat roof 
 rests ; the walls are not structural, but are merely screens 
 enclosing an oblong space within the peristyle ; in the 
 porches they were omitted altogether, as in the porch 
 of the Ashmolean Museum, and in many temples the 
 
22 THE HISTORY OF ARCHITECTURE 
 
 space enclosed by walls was only a very small one in 
 proportion to the area covered by the roof ; there were 
 no windows, and the heads of the doorways were, of course, 
 square, being formed by a lintel spanning the opening 
 from jamb to jamb. 
 
 The working parts i. e. the roof-bearing parts of 
 a Greek building are therefore the entablature and the 
 columns. If the reader will think again of the frame of 
 a billiard-table, he will find that its entablature is formed 
 of three portions ; there is first the edge of a board 
 resting face downwards upon the legs ; above this is the 
 broad plain face of another board set edgeways upon the 
 first ; and projecting over this face is the topmost member 
 in which the openings for the pockets are cut. Similarly, 
 in the entablature of a Greek building there are three 
 divisions : the architrave or lintel proper which rests upon 
 the columns ; the frieze, a broad band of ornament 
 hiding the ends of the rafters; and the cornice or project- 
 ing member which crowns the whole. These divisions 
 will be plain from the illustrations, and may be recognized 
 in the entablature of the Ashmolean Porch or in the 
 Clarendon Building. 
 
 The columns also consist of three parts, the shaft, the 
 base, and the capital. Since they support only a vertical 
 pressure the shafts taper from the base upwards ; they 
 are circular because that form offers least obstruction to 
 the view, and their surface is fluted in order that they may 
 not appear flat against the dazzling sunlight of Greece. 
 
 The flat rooi resting upon the entablature was usually 
 composed of marble slabs. Occasionally it was exposed 
 externally and formed an outdoor platform, but, as a rule, 
 it was protected by an outer framework of timber covered 
 with tiles. 
 
ANCIENT ARCHITECTURE 23 
 
 Three types of columns were used by the Greeks, each 
 carrying its own proper entablature ; as a rule, only one 
 type was used in a single building. All the columns of 
 a particular type conformed strictly to fixed proportions, 
 whether they were large or small ; their capitals were 
 all nearly identical in form and ornamentation ; and all 
 their shafts were channelled with the same number of 
 flutes. Similarly, the entablature belonging to any type 
 was almost indentical in its proportions and ornaments 
 with every other entablature of that type. From a single 
 carved fragment of a Greek building it would be possible 
 to state its order and almost to make a drawing of the 
 whole column and entablature to which it belonged. 
 
 On account of this strict uniformity in the parts of 
 each the three types are known as Orders. They will be 
 described and illustrated in the latter part of the chapter. 
 
 In its lack of variety Classic architecture contrasts 
 strongly with that of the Middle Ages. Ruskin and William 
 Morris held that the stereotyped forms of much of its 
 ornamentation and the almost mechanical accuracy of 
 their execution reflect the fact that Greek architecture, 
 like Greek civilization in general, depended upon slave- 
 labour. Among the ancients, as among ourselves, the 
 design of a master-builder (architekton), worked out in 
 every detail, was executed by workmen whose interest 
 and responsibility were limited to the faithful reproduc- 
 tion of the forms they were set to copy. But every 
 craftsman upon a mediaeval building was in a rude way 
 an artist also, and had his opportunity of leaving upon it 
 the mark of his own individuality. In all the minor details 
 he had a free hand. But the resulting mixture of forms, 
 crude and refined, well cut and ill, grotesque, absurd, 
 graceful, varying with the varying capacities of different 
 
24 THE HISTORY OF ARCHITECTURE 
 
 workmen, would have been foolishness to the Greek as 
 it was to the scholars of the Renaissance. Nevertheless, 
 it is that variety, full of surprise, making its appeal now 
 to the sense of wonder, now to that of beauty, and in 
 the very next carving to that of the ridiculous, that gives 
 to mediaeval ornament its never failing charm. 
 
 THE ORDERS 
 
 The three Orders of Greek architecture were the 
 Doric, the Ionic, and the Corinthian. All three were 
 copied by the Roman architects, who also devised for 
 themselves two other Orders, known as the Tuscan and the 
 Composite. Illustrations are given of the five types, 
 and therefore only a short account of each will be neces- 
 sary. The oldest of the Classic Orders is the Doric ; 
 it perhaps finds its prototype in the columns of the tomb 
 of Beni Hassan in Egypt. It is the most massive of the 
 Orders, its sturdy columns being only about six times 
 greater in height than in diameter, and they are therefore, 
 in themselves, so impressive as to need no base, but they 
 are sometimes mounted on two or three steps. The shaft 
 is fluted with twenty channels, so arranged that a sharp 
 ridge or arris divides each from the next. The abacus 
 of the capital is a plain square slab. The architrave, 
 representing the original wooden lintel, is also plain, 
 and upon it rests the frieze which is ornamented with 
 the characteristic triglyphs, projecting blocks, each 
 channelled with two glyphs or furrows in the middle and 
 with a half-furrow on its side edges. These are considered 
 to have been derived from the projecting ends of the 
 wooden roof-beams once resting at right angles on the 
 
GRECIAN DORIC. 
 
 DORIC. 
 
 FIG. i. DORIC ORDER 
 
26 THE HISTORY OF ARCHITECTURE 
 
 lintel. The spaces between the triglyphs are called 
 metopes, and in the best Greek examples were filled with 
 sculptured figures in relief. The Romans, with the lack 
 of artistic feeling that made them so successful in other 
 directions, substituted oxen's skulls for the figure sculpture 
 they were incapable of copying. Beneath the triglyphs are 
 peg-like ornaments called guttae, supposed to represent 
 the wooden pegs with which the parts of a timber 
 building are held together. Under the cornice, which 
 represents the eaves, is a range of blocks called mutules, 
 suggesting the projecting ends of rafters. 
 
 Although the details of this earliest Greek Order do 
 suggest an origin in timber construction, they must not 
 be regarded as meaningless survivals retained by unin- 
 telligent builders ; the truth lies the other way : details 
 which had become useless structurally were transformed 
 and utilized ornamentally, and they in no way resemble 
 the tentative efforts of a craftsman in wood working in 
 a new material. Some writers even refuse to admit that 
 they are constructive in origin since they are so artistically 
 right for their purpose ; such writers point to the flutes 
 of the shaft which were undoubtedly cut for purely 
 artistic reasons, but which imaginative persons have 
 supposed to be designed for spear-rests and apply the 
 obvious argument to the other details. 
 
 We have no building of the pure Doric form in Oxford, 
 for the Renaissance architects, of course, copied the Roman 
 forms. The Clarendon Building is of the Doric type, 
 and there are Doric columns in the facade of No. 86 
 High Street. 
 
 When the Romans came to use this Order they spoilt 
 it by altering its proportions, lengthening the shaft and 
 narrowing the architrave, and by omitting the sculpture 
 
GREEK 
 IONIC ORDER 
 
 ROMAN 
 IONIC ORDER 
 
 FIG. 2. IONIC ORDER 
 
28 THE HISTORY OF ARCHITECTURE 
 
 and the flutes. Then, dissatisfied with the result, they so 
 modified the original form as to produce what has been 
 named as a new Order, the Tuscan, though it is really 
 a plain form of Doric with the ornaments omitted and 
 a pedestal added to the column. It is represented in 
 modern Oxford in a diminutive form by the columns 
 of the drinking-fountain on S. Clement's Plain. 
 
 The Ionic Order is much lighter than the Doric ; its 
 columns are slender,usually about nine diameters in height, 
 and are widely spaced ; they are fluted with twenty-four 
 channels, separated by flat fillets ; they stand on what 
 is known as the Attic base, consisting of two torus mould- 
 ings, bold rounds, separated by a hollow called a scotia ; 
 an ornament of interlacing circles, called the guilloche, is 
 often carved upon the lower and larger torus. The Ionic 
 capital is distinguished by large volutes at its corners, 
 derived probably from Persian architecture, in which 
 the spiral was always a favourite decorative form. Between 
 the two volutes is carved the echinus or egg and dart 
 ornament, afterwards used profusely by the Renaissance 
 builders in cornices, string-courses, and many other places. 
 The frieze is sometimes plain, sometimes filled with 
 figure sculpture. Beneath the cornice is a row of oblong 
 blocks called dentils ; . they replace the mutules of the 
 older Order. 
 
 The Romans made little use of this Order, and so it 
 changed little in their hands. They reduced the size of 
 the volutes and the number of curls in the spiral, and so 
 spoilt the capital ; they omitted the flutes on the shaft, 
 and increased the projection of the cornice and the size 
 of the' dentils. 
 
 The volute, the most striking feature of the Ionic 
 Order, perhaps of any Order, is the most persistent 
 
n n n n n O 
 
 CREEK 
 CORINTHIAN ORDER 
 
 ROMAN 
 CORINTHIAN ORDER 
 
 FIG. 3. CORINTHIAN ORDER 
 
30 THE HISTORY OF ARCHITECTURE 
 
 ornament in architecture ; old in Egypt, it was copied 
 by the Persians, and adopted by the Greeks in two of 
 their three Orders ; our own forefathers played with it 
 in later times, at first essaying rude copies and then 
 transforming it, as their skill increased, into a knob of 
 opening foliage. At its ultimate origin we can only guess ; 
 but we know that primitive basket-work was coiled, not 
 woven, and that the earliest pottery was similarly formed 
 by strips of clay wound in a spiral round a whorled 
 nucleus. 
 
 Our best example of the use of the Ionic Order in Oxford 
 is in the Taylorian Building in Beaumont Street ; this 
 was built by Cockerell in 1846, and represents the Greek 
 form of the Order. S. Paul's Church has columns of the 
 Ionic type in its western front. 
 
 The Corinthian Order is the last and most graceful 
 invented by the Greeks, and the one most favoured by 
 the Romans and, therefore, by later architects. Its 
 distinguishing beauty is the capital crowned with acanthus 
 leaves, almost the only naturalistic decoration used by 
 the ancients. Elsewhere, distrust of or contempt for the 
 powers of the workman limited him to the execution 
 of simple conventional forms ; but in the main feature 
 of this Order his subject, though conventionally treated, 
 was directly derived from a natural leaf. But the Greeks 
 employed this Order very little. 
 
 The other details of the Corinthian Order resemble 
 those of the Ionic, except that the top of the cornice is 
 ornamented with a row of carved ornaments giving it 
 a broken line. The abacus of the capital is concave at its 
 edges, and its projecting corners are supported by leaves 
 curled into small volutes. 
 
 The Romans made a profuse use of this Order, the 
 
MOULDINGS GREEK AND ROMAN 
 
 FIG. 4. CLASSIC MOULDINGS 
 
32 THE HISTORY OF ARCHITECTURE 
 
 richness of which appealed to their extravagant taste, 
 and by enlarging the volutes and adding the egg and 
 dart ornament to the capitals they produced a form which 
 was afterwards classed as a new Order and named the 
 Composite, as combining the essential features of Ionic 
 and Corinthian. So with the three original Orders and 
 the Tuscan and Composite we have what the Renaissance 
 architects recognized as the ' Five Orders of Architec- 
 ture '. Modern students, however, refuse to consider 
 the Roman modifications as separate Orders. 
 
 The columns of All Saints' Church (1710) and of the 
 Gateway of the University Press (1830) are of the Corin- 
 thian type. In the screens of some of the college chapels 
 and libraries, e. g. in the chapel of Lincoln, columns of 
 this Order may be seen executed in wood. 
 
 Modified forms of all five Orders are placed one above 
 another in the five stories of the western facade of the 
 Schools' Tower (Fig. 79), the sturdy Tuscan forming the 
 base, and then in order the Doric, Ionic, Corinthian, and 
 Composite. The last four may also be seen super- 
 imposed in the tower of Wadham Hall, and the tower 
 of the Fellows' Quadrangle of Merton shows three Orders 
 similarly arranged. All these examples are of the time 
 of James I. In the modern front of Hertford College the 
 Tuscan Order forms the ground floor and the Corinthian 
 the upper story. 
 
 But it is not so much in our classical buildings that the 
 influence of the ancients is to be discerned as in the minor 
 details of buildings of all periods. When one has become 
 familiar with the forms of the Orders, one recognizes 
 their details everywhere. They were rudely imitated 
 by Saxon and Norman, and beautifully modified by 
 Gothic workmen ; they were slavishly copied by the 
 
ANCIENT ARCHITECTURE 33 
 
 Renaissance builders, so that even the tombstones in the 
 churchyards are carved with pagan ornament ; and in 
 our own day the mouldings of our mantelpieces, our 
 door-panels, and window-sashes, even of our picture- 
 frames, are usually debasements of Classic forms ; the 
 very lamp-posts in our streets are derived from the 
 Classic column. All roads in architecture lead at last 
 to Rome. 
 
 CHAPTER II 
 
 THE INVENTION OF THE ARCH 
 
 IN the middle of the second century B.C. the Romans 
 conquered Greece, and having no national architectural 
 style of their own, for they were apparently still in the 
 timber stage of building evolution, they adopted the Greek 
 Orders, and, importing Greek architects, began to make 
 Rome a city worthy of the empire of which she was 
 the mother. 
 
 But they were soon confronted with the difficulties 
 inherent in the lintel system. Theirs was a complex 
 civilization, and they needed buildings of many types ; 
 great covered spaces for public purposes, for instance, 
 were as necessary in Rome as in London now. But the 
 multitude of close-set columns obstructed sound and sight 
 besides occupying valuable space ; how to decrease the 
 necessary points of support for the roof was the first 
 problem for their architects. A complete account of the 
 solution of that problem would be the history of building 
 construction from the first century to the fifteenth. 
 Again, the consideration that has in our time produced 
 
 785 C 
 
34 THE HISTORY OF ARCHITECTURE 
 
 the American ' sky-scraper ', the value of surface-area in 
 congested sites, was beginning to present itself in populous 
 Rome ; and so there was the further problem, how to 
 construct a building of several stories, for column and 
 lintel could bear no great weight of superstructure. 
 And lastly, there was the question of material ; the diffi- 
 culties of obtaining an adequate supply of large stones 
 for the amount of building planned and in progress were 
 enormous. How could material of a more ordinary kind 
 be used instead of the column and lintel ? The Roman 
 architects triumphantly overcame all these difficulties 
 by using arches instead of lintels in their buildings. 
 
 The arch was no new invention ; it had been used by 
 the Etruscans even before the Romans became rulers in 
 Italy, and the latter must therefore have been quite 
 familiar with its form. But no nation had yet attempted 
 to make it the base of an architectural system, for none 
 had yet been driven by new conditions to seek a sub- 
 stitute for the lintel. Moreover, while the arch solved 
 one problem it raised another, for, unlike the lintel, it 
 could not be carried on columns ; it tended to thrust 
 apart its supports and so to fall with them into hopeless 
 ruins. This must have been very discouraging to the 
 early experimentalists ; hence we find that practically 
 all early arches are used merely to cover drains and sewers, 
 where they carry little weight and are prevented from 
 spreading by the sides of the trench in which they are 
 built ; the solid mass which serves this purpose is called 
 the abutment of the arch : without it no arch can stand ; 
 and the Romans were the first people to understand clearly 
 this fundamental fact in the mechanics of the arch, and 
 to realize that, given sufficient abutment, there is prac- 
 tically no limit to its weight-bearing capacity. The 
 
THE INVENTION OF THE ARCH 35 
 
 discovery of the arch as a weight-carrier, as the very 
 Atlas of architecture, and the consequent substitution of 
 the arch for the lintel, is the most important step in the 
 history of architecture since first man laid one stone 
 upon another to make himself a shelter. 
 
 Stones used in an arch are in a state of compression, 
 i.e. the material is resisting just that force which "its 
 granular nature is formed to resist ; stones of all sizes can 
 be so used ; and arched construction, unlike the lintel 
 system, is therefore possible in every district where stone 
 exists at all. After thousands of years of stone building 
 it remained for the Romans to show mankind the true 
 use of the material, and to prove that the natural basis 
 of stone architecture is the arch, as that of timber con- 
 struction is the beam. 
 
 But the Romans had at first no idea of inventing a new 
 system ; they admired too much the grandeur of Greek 
 building, and had no thought of abandoning the horizontal 
 line that gave it its repose and dignity. What they desired 
 was to support an entablature without columns, and to 
 the last they did not see that the idea was an absurdity, 
 since a lintel supported along its whole length ceases 
 to be a lintel at all. 
 
 The simple illustration of the billiard-table will help 
 the reader to understand the new principle. It is required 
 to get rid of the intermediate legs and to support the 
 oblong 'entablature' without them : now if an arch is 
 built between the end legs of each side, the middle point 
 of the entablature will rest upon its crown ; further, if 
 a wall is raised upon the haunches of the arch to the level 
 of the crown a continuous support will be formed, making 
 it possible to construct the entablature in short lengths ; 
 similarly, other arches built on the other sides of the oblong 
 c 2 
 
36 THE HISTORY OF ARCHITECTURE 
 
 will carry the remaining sides of the entablature. But 
 the four arches will carry the slate bed without the need 
 for an entablature ; moreover, they render superfluous the 
 legs at the corners, and so lintel and column can alike 
 be dispensed with. Our illustration supposes an arch 
 springing from the ground-level and having its abutment 
 in 'foundations below. But practically, of course, the 
 entablature carries a roof, and therefore the supporting 
 arch must spring from a level above the ground. Now, 
 the Greek column, being either a monolith or else 
 constructed of drums, had been formed to carry a vertical 
 weight and not to resist the oblique thrust of an arch ; 
 a much greater mass was necessary to provide the abut- 
 ment for this purpose. The Roman architect, therefore, 
 replaced the column by a solid mass of masonry (called 
 a pier), and built his arch upon this. But though the 
 column was thus rendered useless, he would not abandon it, 
 but attached it as a pilaster to the face of the pier (Fig. 5), 
 and carried it up to the level of the entablature, making 
 it appear as if it still performed its ancient function and 
 supported an architrave that really rested upon the arch. 
 Its real structural purpose in this position was to weight 
 the pier and so provide greater abutment, but it seemed 
 not to do this, and to do something else which in reality 
 it did not do ; it was therefore a constructive lie and 
 a blemish upon the Roman system. The excuse for the 
 Roman architects is that they could not realize their inven- 
 tion to be revolutionary ; no such excuse can be made 
 for the Renaissance builders, who in their blind admiration 
 for the ancients copied the falsehood ; still less can the 
 modern imitator be excused, he is like the Hindoo shoe- 
 maker who copies the slits and patches of the boot 
 given him for a model. 
 
THE INVENTION OF THE ARCH 
 
 37 
 
 Though to the last he tried to disguise the fact, the 
 Roman architect completely abolished the old trabeated 
 
 FIG. 5. SHELDONIAN THEATRE 
 
 system and developed to the full the new arcuated style ; 
 he might hide it in the facade of his building, but in 
 the essential structure, the roof, the triumph of the arch 
 
38 THE HISTORY OF ARCHITECTURE 
 
 was clear and unmistakable : there was a revolution in 
 architecture. 
 
 Where roofs were of wood they continued to be flat 
 on their under sides, being, of course, still formed of 
 transverse beams ; but they were carried not by posts 
 or columns, but by arches ; timber-roofed buildings 
 therefore involved a mixed system. 
 
 But the thoroughness that was the vital characteristic 
 of the Roman mind would not allow their builders to 
 
 FIG. 6. BARREL VAULT 
 
 rest here. They had discovered how to support a roof 
 without columns ; they now proceeded to discover how 
 the architrave and beams might be dispensed with and 
 a building constructed on the arcuated system throughout. 
 That is, they invented arched roofs of stone, called vaults. 
 The earliest form of vault followed inevitably upon the 
 invention of the arch ; in a sense every covering of a 
 vertical opening the head of a doorway or a window 
 for instance is a roof, and any archway, however narrow, 
 is thus a vault. An arch of Magdalen Bridge, for example, 
 is an arch as seen from without, but a vault to a person 
 beneath it. It is obvious that when once an arch had been 
 constructed, the idea of increasing its depth longitudinally 
 
THE INVENTION OF THE ARCH 
 
 39 
 
 to form a roof would immediately follow. So was pro- 
 duced the first stone vault, known from its form as the 
 tunnel or barrel vault. Examples of this type are found 
 at all subsequent dates ; in Oxford a barrel vault of the 
 twelfth century roofs the slype of the Cathedral Cloisters, 
 and another, of the Renaissance period, the kitchen of 
 Wadham. 
 
 Any rectangular area could be roofed by this means, 
 but there were two disadvantages involved ; abutment 
 was necessary along the whole length of the vault, neces- 
 
 FIG. 7. GROINED CROSS-VAULT 
 
 sitating very massive supporting walls ; and all openings 
 for doors and windows must be cut below the level of the 
 springing of the vault. 
 
 Now, if, as often must have happened, one such 
 vaulted area, as for instance a gallery or passage, was 
 crossed by another at right angles, what complications 
 would ensue ? If we imagine two equal tunnels to inter- 
 penetrate, no part of either being omitted, the square 
 formed by the crossing of their rectangular plans will be 
 enclosed within four walls, blocking the passage-way, and 
 will be roofed by parts of the two vaults above it. The 
 intersection of these vaults forms diagonal arches called 
 
40 THE HISTORY OF ARCHITECTURE 
 
 groins, which spring from the corners of the square and 
 are independent of the support of its walls ; the drawing 
 shows that if these walls and parts of the two vaults are 
 removed, the square space will be roofed by four curved 
 stone faces supported by the arches of the groins (Fig. 7). 
 
 The construction of this cross-vault, though difficult to 
 explain in words, is perfectly simple to the eye when seen 
 in a drawing or a model. A clear idea of it is essential 
 to the appreciation of all later architecture, for its dis- 
 covery by the Romans was the beginning of that system 
 of the concentration of arch-thrusts which is the funda- 
 mental principle of Gothic building construction. When 
 once it was discovered that a square space could be 
 roofed by a groined vault, resting on the four angles 
 only, it was seen that any rectangular area could be 
 vaulted by dividing it into squares and roofing each with 
 a cross-vault. The walls could then be omitted or reduced 
 to mere screens, provided that piers were left at the corners 
 of the squares to support the springing of the groins. 
 
 The Romans vaulted enormous areas with single cross- 
 vaults ; no later architects dared to emulate them. 
 The early mediaeval workmen divided their spaces into 
 squares, and roofed each with a small groined vault ; but 
 they learned to improve upon their model, as will pre- 
 sently be seen, and they devised a new and superior system. 
 The Renaissance architects put back the hands of the 
 clock and returned again to the simple groined form that 
 had been improved out of existence. Examples of the 
 revival of Roman cross-vaulting may be seen in the work 
 of the eighteenth-century architects, in the cloisters of 
 Queen's, All Souls, and Worcester Colleges. 
 
 The immense importance of this Roman legacy to 
 posterity, the groined vault, will become more apparent 
 
THE INVENTION OF THE ARCH 41 
 
 when we come to consider our Norman buildings : still 
 more when we try to trace the progress of roof-making 
 from the simple barrel vault of the Cathedral slype to 
 the intricate mechanism of the vaulting of the choir 
 
 That arching roof, 
 
 Self poised, and scooped into ten thousand cells 
 Where light and shade repose, where music dwells 
 Lingering and wandering on, as loath to die. 
 
 The groined vault was the germ from which the 
 peoples of Western Europe developed the marvellous 
 'fretted vaults' which are the most characteristic feature 
 of Gothic ; this application of the arch was Rome's 
 bequest to her western heirs. But so magnificent was 
 her estate that she had an almost equal gift for the 
 eastern nations ; it was that noble form of arched roof 
 that causes the Radcliffe Camera to dominate every view 
 of Oxford and makes S. Paul's Cathedral seem to group 
 all London around its mighty dome. 
 
 The dome, like the barrel vault, followed inevitably 
 upon the discovery of the arch ; its form is produced by 
 the revolution of a semicircle upon its vertical axis. 
 But though the Romans were the first builders of great 
 domes, as they were of scientifically-buttressed arches, 
 they were anticipated in the use of that form of roof 
 by prehistoric man in Egypt, whose pit-dwellings were 
 covered by domes of dried mud, and, ages earlier still, 
 by the beavers, which roofed their circular lodges with 
 domical vaults of twigs and clay. 
 
 The obvious application of the dome is to the roofing 
 of a circular space ; the Romans, having invented the 
 form and applied it to round buildings, left it to their 
 successors in the eastern empire to poise it upon the 
 
42 THE HISTORY OF ARCHITECTURE 
 
 angles of a square and abut it by lesser domes or barrel 
 vaults, thus giving the Byzantine architects the chance 
 to win immortality for their memories by their use and 
 development of their Roman heritage. 
 
 In Oxford we have the dome as the Romans left it, 
 for the Renaissance architects, to whom all our examples 
 are due, of course copied the buildings of Rome. It is 
 only in details that our English styles show Byzantine 
 influence, and therefore, in a book of local architecture, 
 we are not justified in tracing the development of a 
 divergent branch; but to those interested in architecture 
 as a whole and to this state all must come who develop 
 any real interest the adaptations of the dome by the 
 Byzantine architects will prove almost as interesting as 
 the development of cross-vaulting by those of the Gothic 
 races. 
 
 One other application of the arch remains to be con- 
 sidered, its use in the building of bridges. There had 
 been great buildings before its invention, but there had 
 been and there could be no great bridges. The influence 
 of bridges upon civilization has been incalculable, and 
 the use of the arch in carrying highways over rivers is 
 probably the highest service its invention has rendered 
 to the welfare of humanity. 
 
 CHAPTER III 
 
 ROMANESQUE ARCHITECTURE 
 
 MODERN architecture, like modern history, begins where 
 Roman ends, with the wave of barbarism that over- 
 whelmed the empire. And like the history of modern 
 civilization, its story is the tale of the reconstruction, 
 under Christian influence, of the old ruins into a nobler 
 
ROMANESQUE ARCHITECTURE 43 
 
 system than the ancient one. So great was the shock, 
 that centuries passed before the arts began again to lift 
 their heads, and during those Dark Ages Western Europe 
 relapsed into timber construction. Only in the East, and 
 especially in Byzantium, art and learning still survived. 
 In Italy also there was a sort of continuity in architec- 
 ture, the Christian barbarians building, or rather concoct- 
 ing, churches, by piecing together fragments of Roman 
 buildings 
 
 In England the palaces of the kings, and the cathedrals 
 of the bishops, were of such unsubstantial character that 
 scarcely a vestige remains that may with certainty be 
 ascribed to an earlier date than the eleventh century. 
 Yet even in the Dark Ages, and in England, the darkest of 
 European countries, the influence of Roman architecture 
 was never without its witness. Monks returning from 
 pilgrimage to Rome kept alive the tradition of stone 
 buildings, and Bede records attempts to construct 
 churches ' with stone and lime ', ' after the Roman 
 manner.' Traces of these early Saxon churches remain 
 in the crypts of Ripon and Hexham, and nearer home at 
 Wing, Barnack, and Brixworth the last being possibly 
 a reconstruction of a Roman basilica. 
 
 The history of Oxford begins in the eighth century, with 
 the story of S. Frideswide, and one would like to believe 
 that our architectural history begins there too, with the 
 building of the church of her nunnery. In the east wall 
 of what is now the Lady Chapel of the Cathedral are 
 three round arches of rude workmanship that must 
 certainly belong to an earlier building ; and on the other 
 side of the wall, in the Canons' Gardens, are buried the 
 foundations of apses into which these arches opened. 
 The workmanship of the arches is just what might be 
 
44 THE HISTORY OF ARCHITECTURE 
 
 expected from men making a rudimentary attempt to 
 build on the Roman model, with no other guide than 
 rough sketches and the oral instructions of some returned 
 pilgrim. The stones are unhewn, merely the local rubble 
 in lumps, just used as it came straight from the quarry, 
 and their irregularities make thick beds of mortar neces- 
 sary to fit them together. It is then not improbable that 
 we have here the traces of an eighth-century church, and 
 an example of the early rude attempts at the revival 
 of Roman architecture such as are recorded by Bede. 
 
 But it was not until the genius of Charlemagne had 
 reduced the chaos of the western empire into some sort 
 of order that the nations had leisure and quiet enough 
 to make any general attempt to evolve an architectural 
 style. Then as the tumult and the shouting dies, white 
 churches begin to be dotted over Germany and France, 
 each fondly believed by its builders to represent the real 
 Roman manner. The East has long ago developed for 
 itself a Romanesque style of building, and now a western 
 form of Romanesque architecture comes also into being. 
 This, if it were possible 'at this time of day to correct 
 the nomenclature of the Renaissance architects, is the 
 true Gothic style the building system evolved by the 
 northern races of whom the Goths are representative. 
 
 In England its progress was slow, for hardly had Angle 
 ceased to slaughter Saxon than Dane arrived to massacre 
 both. Britain also was farthest removed from Rome, 
 was the last part of the Empire to be civilized and the 
 first to be abandoned to the barbarians, and so there 
 remained but few Roman buildings to serve as models or 
 as quarries of hewn stone. One would therefore expect 
 to find few examples of stone buildings belonging to the 
 Saxon period ; the rude arches in the Cathedral are 
 
FIG. 8. SAXON ARCH IN CATHEDRAL (c. 727, part restored) 
 
46 THE HISTORY OF ARCHITECTURE 
 
 
 
 all that remains of early Saxon Oxford except the great 
 mound (c. 912) that formed the fortress of the citizens. 
 
 Churches there must have been ; for though the first 
 evidence of the existence of Oxford does not occur until 
 the year 912, yet the entry in the Chronicle under that 
 date shows that Oxford was already an important town 
 and the military centre of the surrounding district. ' In 
 this year died ^Ethered ealdorman of the Mercians and 
 King Eadward took possession of London and of Oxford, 
 and of all the lands that owed obedience thereto.' A town 
 classed with London as a military centre must have had 
 a considerable population and a corresponding number of 
 churches. But of these not a vestige remains. Either 
 they were of wood or else of such primitive masonry that 
 they were contemptuously swept away by the Norman 
 builders. Oxford in Saxon days, a border town now 
 in Wessex, anon in Mercia, alternately burnt by either side 
 and by the Danes when Wessex had permanently secured 
 it, must have been a most unhappy city in which church- 
 building was a matter not likely to be very much in the 
 minds of the population. The Danes, who had no regard 
 for the Sabbath, welcomed an opportunity of catching 
 folk at their prayers and burning their church over their 
 heads. Not until the eleventh century had Oxford 
 people any reasonable certainty of undisturbed devotions. 
 
 Alien sovereigns have ever been blessings in disguise 
 to England, and the accession of Canute not only gave 
 peace to a harassed land, but brought its people into 
 touch with European movements. So now, in the middle 
 of the eleventh century, Romanesque architecture, hitherto 
 represented only in a few favoured situations, spread 
 over the country and left its mark on Oxford in the 
 building of the tower of S. Michael's Church. A very 
 
FIG. 10. S. MICHAEL'S TOWER (c. 1071) 
 
ROMANESQUE ARCHITECTURE 49 
 
 human wish to associate a piece of work with a name 
 has caused the tower to be attributed to Robert d'Oilgi. 
 There is no evidence either way, and it is quite immaterial 
 whether the tower was built shortly before or shortly 
 after the Conquest. What is certain is that it exhibits 
 all the characteristics of buildings erected by the English 
 before the influence of the more skilful Norman masons 
 had affected the native craftsmen. For a generation 
 after the Conquest the Normans must have been too busy 
 with military matters to think of church architecture, 
 and such churches as were built must have been con- 
 structed by native workmen very much in the pre-Conquest 
 manner. 
 
 An entry in Domesday Book states that the ' Priests 
 of S. Michael ' had two mansions in Oxford at the date 
 of the survey ; the Chronicle of Abingdon Abbey records 
 that Robert D'Oilgi restored churches both within and 
 without Oxford ; and the Oseney Chronicle says that 
 he built S. George's Church in the Castle and endowed it 
 with lands for the support of its priests. That is all the 
 documentary evidence, and all it proves is that D'Oilgi 
 was a church-builder, and that S. Michael's Church existed 
 in his day. The latter fact is also proved by the architec- 
 ture of the tower ; and one detail, the moulding of the 
 impost stones of its belfry windows, suggests that one 
 of his Norman masons may have taken part in building 
 or restoring it. 
 
 The rest of the church shows a mixture of the work 
 of different dates in which no details are earlier than the 
 thirteenth century. Judging by the lancet windows of 
 the east end, this is the date of the present chancel ; 
 there is, however, one very remarkable feature which 
 affords some reason for thinking that the walls, in spite 
 
 785 D 
 
50 THE HISTORY OF ARCHITECTURE 
 
 of their thirteenth-century windows, may be those of the 
 original building : their great height compared with 
 the narrowness of the chancel is without parallel in the 
 district ; but any one that has seen the Saxon churches 
 at Bradford-on-Avon and Jarrow will at once be struck 
 by the strong similarity between their high and narrow- 
 structure and that of S. Michael's chancel. 
 
 With the astonishing aptitude for civilization that 
 distinguished them, the Normans had eagerly taken part 
 in the revival of architecture, and at the date of the 
 Conquest were developing in their duchy a local school 
 of Romanesque, known as the Norman style. Its influence 
 on English building is first seen at Westminster in the 
 work of Norman masons imported by the Confessor. 
 At this time Norman architecture differed from Saxon 
 mainly in being on a grander scale, but, transplanted 
 into England, it developed with extraordinary vigour in 
 the century following the death of the Conqueror, 
 and, unlike every other style in history, it knew no decline, 
 and only ceased to exist because it lost its identity by 
 its very growth, as the larva ceases to be a grub by becom- 
 ing a dragon-fly. 
 
 For this reason one ought to judge carefully whether 
 the Norman, or indeed the western Romanesque style as 
 a whole (for Norman is but the most fruitful branch of 
 the western Romanesque tree), is to be considered as 
 a separate style. In the chapter on roof-making I have 
 followed Ruskin in admitting three possible architectural 
 systems, distinguished by the use of the lintel, the round- 
 arched and the pointed vault ; but there is such affinity 
 between the two last that it might be asserted that there 
 are only two great architectural styles, the trabeated 
 system of the ancients and the arcuated system of the 
 
FIG. ii. IFFLEY CHURCH, B \VEST FRONT (c. 1160) 
 
$2 THE HISTORY OF ARCHITECTURE 
 
 moderns. 1 Then Romanesque and Gothic must be 
 regarded as stages of growth in a great arcuated system 
 invented by the Romans and perfected by the mediaeval 
 builders, and for which a name has yet to be invented. 
 
 Englishmen of Norman days saw greater activity and 
 progress in building than has been seen before or since 
 on the face of the earth. The genius of that extraordinary 
 race has its witness in the architecture of three-quarters 
 of the ancient churches of the country. Buildings that 
 they planned for mere monastic churches serve to-day 
 as cathedrals for great bishoprics, some, like our own 
 Cathedral, shorn of their original dimensions. Village 
 churches, such as Iffley, still suffice for the needs of the 
 district, though built when the population of the whole 
 Norman dominions was less than that of modern London. 
 
 In Oxford evidence of this astonishing energy may be 
 found in the tower, the crypt, and well-chamber of the 
 Castle, and the chancel arch of Holywell, all attributed 
 without improbability to the d'Oilgis, in the chancel 
 and crypt of S. Peter's and the whole of its south wall, 
 including the doorway, in the doorway of the Cathedral 
 Chapter House, and in the adjoining slype, in the remains 
 of a doorway at S. Ebbe's, in the basement courses of 
 S. Mary Magdalene tower, in the lately discovered crypt 
 in Frewin Court, and in the nave, choir, and transepts 
 of the Cathedral, which is the last important building 
 in England that can be said to belong to the Norman 
 style. These relics still remain to tell of building and 
 rebuilding in Norman times ; but they represent only 
 a fraction of the results of Norman activity in Oxford ; 
 the Chapter House doorway implies a complete building ; 
 
 1 The mixed system of our own day is, of course, a hybrid and 
 no true style. 
 
54 THE HISTORY OF ARCHITECTURE 
 
 so with the other features. And in addition we have 
 documentary evidence of the building of a complete 
 monastery, Osney, with a church larger than the present 
 Cathedral. 
 
 It was not only in church-building and in fortification 
 that the Normans proved themselves the instruments 
 of civilization ; the greatest of d'Oilgi's architectural 
 achievements in Oxford was the building of his two 
 bridges across the Thames, the Grand Pont * and the 
 Hythe Bridge. Both have been rebuilt more than once, 
 but of his bridge over the Cherwell at Heyford some 
 parts are original to-day. 
 
 All through the Romanesque period successive genera- 
 tions of builders were working out the problems of arch 
 construction that had already been solved by the Romans. 
 
 Their great ambition was to construct a church all of 
 stone from floor to ceiling, but the difficulty of vaulting 
 the nave by means of round arches alone baffled them 
 until the last. They could arch the heads of doors and 
 windows and construct barrel vaults and semi-domes ; 
 later on they successfully attempted the cross-vault in 
 the roofing of aisles and chancels ; but at Durham alone 
 the nave was vaulted by this means. Elsewhere the 
 solution was not reached until the search for it had 
 brought about so many developments and modifications 
 that the Norman style was transformed out of existence. 
 So in our own Cathedral the Norman builders covered 
 the aisles with stone vaults, but did not venture to arch 
 the ceilings of the nave or the choir. That they hoped 
 to be able to do so eventually is shown by the shafts they 
 built to carry the vault-arches, which shafts were actually 
 used for that purpose when the choir-vault was con- 
 1 This is uncertain, but not improbable. 
 
ROMANESQUE ARCHITECTURE 55 
 
 structed in the fifteenth century. But the nave, like most 
 Norman naves, remains incomplete to this day, being 
 still covered by a timber roof. 
 
 The progress of architecture through the twelfth 
 century is almost entirely due to the efforts of the archi- 
 tects to construct a completely vaulted building ; and it 
 was the achievement of their ambition that brought the 
 Romanesque style to an end, and in so doing brought 
 into being the system known to us as Gothic architecture. 
 
 My purpose in this part of the book being only to sketch 
 briefly the history of the successive styles, I shall defer to 
 the second part a detailed description of each ; the reader 
 will already have understood that the use of one or 
 other of the ancient Orders marks a building in the 
 Classic style, that a building with round arches, without 
 columns and entablature, is usually of Romanesque 
 architecture, and that the pointed arch is the most easily 
 recognized characteristic of Gothic. 
 
 CHAPTER IV 
 
 GOTHIC ARCHITECTURE 
 
 THE difficulty of constructing a cross-vault over the 
 oblong bays of the nave by means of round arches led, 
 in the Late Norman period, to the introduction of the 
 pointed arch. It was then found that by using arches of 
 varying span, all rising to the same level, it was possible 
 to support a vault upon many small arches instead of 
 upon the two diagonals of a cross-vault. The result was 
 not only the simplification of vaulting problems, but the 
 discovery of a system of concentrating the thrusts of the 
 roof-arches at a few points where the resultant of their 
 forces could be received by a solid mass of masonry. 
 
56 THE HISTORY OF ARCHITECTURE 
 
 The Norman plan, like the Roman, had been to ensure 
 stability by making the walls and piers so massive that 
 their weight was sufficient to resist the thrusts exerted 
 at any point. But this, except in the building of barrel 
 vaults, involved a great waste of material, since the thrusts 
 of a cross-vault are received at four points only (Fig. 7) ; 
 neither the Roman nor the Norman builder fully realized 
 that at these points alone was abutment needed. 
 
 This was the great conception of the Gothic architects, 
 and the central principle of their system is less the use 
 of pointed arches than the economizing of material by 
 the skilful distribution of masses at the points where 
 arch-thrusts were concentrated and by the disposition 
 of arches so that the thrust of one was met and annihi- 
 lated by that of another. This principle, which will be 
 illustrated fully in another chapter, was conceived in the 
 latter part of the twelfth century, and was worked out 
 through the three centuries following. Its development 
 was a joint enterprise common to the architects of 
 Western Europe ; an advance in one country was almost 
 as quickly taken advantage of in the others as an improve- 
 ment in battleship construction in our own day. 
 
 The new system, unhappily named Gothic by the 
 Renaissance architects, was, still more unhappily, divided 
 into three styles by the nineteenth-century Gothic 
 revivalists. When Rickman, the earliest of them, came 
 to examine the characteristics of the system he, like the 
 pioneers in every science, classified his subject accord- 
 ing to details, without appreciating the principles that 
 underlay them. Observing that the superficial character 
 of buildings of the thirteenth century differed con- 
 siderably from those of the fourteenth, and these again 
 from those of the fifteenth, he classified the thirteenth- 
 
GOTHIC ARCHITECTURE 57 
 
 century work as Early English, that of the fourteenth as 
 Decorated, from the more ornate character of its detail, 
 and that of the fifteenth as Perpendicular, from the 
 prevalence of the right angle in its window-tracery and 
 panelling. But in truth, there are no more three 
 styles in Gothic architecture than there are three 
 persons in one individual. What he mistook for styles 
 are stages of development. These names, however, like 
 the unfortunate word Gothic, are now fully established, 
 and it seems hopeless to try to abolish them. But they 
 are certainly misleading, and they are incomplete even 
 as representing stages of development, so that ' transition ' 
 styles have been invented to describe the work of the 
 end of the thirteenth and fourteenth centuries. This is 
 to make confusion worse confounded, for it is to assert 
 definite periods of transition while the essential truth 
 about Gothic architecture is that it was in transition, 
 i. e. developing, throughout the whole of its existence. 
 It is, of course, possible to distinguish broad stages in 
 the growth of Gothic as in that of a human organism ; 
 an alternative to Rickman's classification is that of 
 Sharpe, who recognizes the following periods : 
 
 Lancet, 1180-1245 : example Cathedral Chapter 
 
 House, 1 220. 
 
 Geometric, 1245-1315: example Merton Chapel, 1297. 
 Curvilinear, 1315-60: example Latin Chapel, 1350. 
 Rectilinear, 1360-1550 : example Divinity School, 
 1450. 
 
 But these, though more complete than Rickman's divisions, 
 are based, like his, upon superficial characteristics ; and 
 they are even less descriptive, because they refer merely 
 to one conspicuous feature the window. They are no 
 
58 THE HISTORY OF ARCHITECTURE 
 
 more a scientific division of stages than Shakespeare's 
 picturesque Seven Ages of Man is descriptive of the 
 organic growth of the individual man. A really logical 
 distinction of the periods of Gothic architecture would 
 be based upon the progressive changes in vault construc- 
 tion and buttressing, to which modifications in other 
 features can usually be traced. 
 
 It may be objected that the great majority of Gothic 
 buildings are not vaulted at all. The answer is that they 
 are not representative Gothic, but only buildings with 
 Gothic details. Gothic construction was worked out in 
 the great monastic architectural schools, and its details 
 were copied in the parish churches whether they were 
 structurally necessary or not. The parish churches reflect 
 the progress of Gothic, but they do little to ex-plain it. 
 
 A description of Gothic architecture can be illustrated 
 by the details of almost any ancient church, but the 
 morphology of those details can only be properly under- 
 stood from an examination of a vaulted building. Few 
 students in these days will be content to know merely 
 the forms by which the work of different periods is 
 characterized ; in the study of architecture, as in that 
 of science, to know the forms of things is only to ask 
 how those forms are adapted to their purpose, and by 
 what stages they reached their existing shapes. 
 
 The pointed arch, which had been found essential in 
 the vaulting, was soon discovered to be graceful and con- 
 venient for use in the other parts of the building. At 
 first the architects used the new form only where it was 
 structurally more suitable, but its beauty, and perhaps 
 the desire for consistency, soon led to the almost total 
 exclusion of the semicircular arch. 
 
 We have in our own Cathedral a building that illustrates 
 
60 THE HISTORY OF ARCHITECTURE 
 
 as well as any in England the passing of Romanesque 
 architecture into Gothic. The chancel was begun in 
 1160, and the west end of the nave was finished in 1180. 
 Richard I, who lived as a child in Beaumont Palace, may 
 well have been taken to see the builders at work upon 
 the new church of S. Frideswide. If he had spent his 
 boyhood's years in Oxford he might have seen the gradual 
 change of character and method that marked the slow 
 progress of the work. East of the tower every arch is 
 semicircular, the piers are ponderous, and the vault-ribs 
 plain and massive ; but in the nave, though the main 
 arches are round, the heads of the windows above are 
 pointed, and so are the arches of the aisle-vaults, the 
 piers are less massive, and the vaulting-ribs are light and 
 thin, and their surface is relieved by grooves and ridges 
 (mouldings) carved in the stone. 
 
 The workmen that finished the west end of the Cathedral 
 may have found work awaiting them in the building of 
 S. Giles's Church. There is very little at S. Giles's that 
 is later than the middle of the thirteenth century, and the 
 tower-arches and south aisle belong at the latest to the 
 earliest years of that century. The wall of that aisle 
 is still thick and massive, for it has but two small 
 buttresses at its angles ; its windows, set singly in the 
 wall, have the long thin shape that marks the Lancet 
 or Early English stage of Gothic ; and on the con- 
 temporary font may be seen the ' dog-tooth ', which is 
 the characteristic ornament of the style. The builders 
 of the north aisle may have discussed the judgement 
 of heaven upon John as they sat at their midday meal. 
 The lancet windows are now grouped together in twos 
 and threes, and the work is much lighter than before. 
 The walls are well buttressed, and are therefore thinner 
 than the northern one. 
 
62 THE HISTORY OF ARCHITECTURE 
 
 By far the finest example of Early Gothic in Oxford 
 is the Chapter House of S. Frideswide, built about 1220. 
 Six pointed arches carry the vaulted roof of each of its 
 bays ; it has single lancet windows in its side walls 
 and a group of five at its east end ; the deeply-cut 
 mouldings that ornament their arched heads, and the 
 foliage that is cut upon the caps of their shafts are admir- 
 ably typical of the best work of their date. The contem- 
 porary Lady Chapel of the Cathedral is also a fine example 
 of the Lancet or Early English stage of the style. 
 
 Of its next stage, the Decorated, in which it reached 
 its highest beauty, we have one of the most beautiful 
 examples in the choir of Merton Chapel, built in the last 
 years of the thirteenth century. Unfortunately it is not 
 vaulted, and to examine the development of vaulting 
 science we must go to the Latin Chapel of the Cathedral, 
 half a century later in date ; but in its windows and its 
 carved ornament it shows as well as any building in 
 England the perfection of Gothic art. The Lancet stage 
 is definitely left behind, and instead of separate lancet- 
 shaped piercings of the walls, the various openings are all 
 formed by means of a framework of stone bars inserted 
 in a single large arch ; the lights are lancet-shaped still, 
 but they form only a minor part of the composition ; 
 it is the design above them that is the conspicuous 
 feature in the windows of this period, the tracery of circles 
 and trefoils, curved triangles, and other geometrical 
 forms, that fills the arched head of the window with 
 beautifully shaped openings for the admission of light. 
 
 In the canopies of the stone seats near the altar, and 
 in the carved foliage of the contemporary shrine of 
 S. Frideswide in the Cathedral, we have some of the 
 loveliest examples of mediaeval art. For the first time in 
 
GOTHIC ARCHITECTURE 65 
 
 the history of architecture the beauty of natural leafage 
 is faithfully reproduced in the decoration of buildings. 
 But this period is all too short ; S. Mary's spire (c. 1300) 
 is its highest achievement ; then Gothic art begins its 
 slow decline, though Gothic science is progressive to the 
 end. The artistic beauties of the style belong to its 
 early stages ; its mechanical triumphs to its later years. 
 
 Of the century between the dates of Merton and New 
 College we have few buildings remaining. The windows 
 of the south aisle of S. Mary Magdalene, of the north 
 aisle of S . Peter's, and of the Latin Chapel in the Cathedral, 
 all belonging to the first half of the fourteenth century, 
 show in their tracery the wavy lines of the ' Curvilinear ' 
 or late Decorated period. The buttresses of these 
 buildings, with their increased projection and the con- 
 comitant thinness of walls, are also representative of their 
 date. But we have no piers or doorways of the mid-Gothic 
 period, and our only fourteenth-century vault 1 is that 
 of the Latin Chapel. It is not so elaborate as many 
 others of its date, but it will serve to show the progress 
 of vaulting in the century following the building of the 
 adjacent Lady Chapel. 
 
 The Black Death put a check on the development of 
 architecture for nearly a generation. In 1380 William 
 of Wykeham began his ' new ' college, and adopted 
 a new form of window-tracery which had been invented 
 at Gloucester. In these windows the stone bars between 
 the lights are carried up from the sill to the arch, thus 
 giving support to the latter, and allowing it to be made 
 wider. But mullions of such a length require lateral 
 stays, and so transoms, or cross-bars of stone, were carried 
 
 1 Except the plain vaults in the basement of the Old Convocation 
 House (c. 1320) and in the passage-way to the Mob Quad at Merton. 
 785 E 
 
66 THE HISTORY OF ARCHITECTURE 
 
 horizontally across the window from jamb to jamb. It 
 was these great windows, with their numerous oblongs 
 formed by crossing mullions and transoms, that suggested 
 to Rickman the term Perpendicular to describe the 
 fifteenth-century stage of Gothic. The rectangular 
 forms in the windows are repeated in the panels cut in 
 the face of the walls ; these, though they offend artisti- 
 cally by repeating a single form, are an evidence that the 
 great Gothic principle of economy in material was never 
 better applied than in the last stage of the style; a panelled 
 wall is as efficient as a blank one, and requires less stone. 
 But it is in the vaulted roofs of the fifteenth century 
 that the triumph of Gothic science is most clearly seen. 
 The vault of the Divinity School, for example, is a 
 mechanical marvel ; it is supported by means of buttresses 
 alone, the walls between them being practically sheets of 
 glass. And even in the buttresses, great as they are, not 
 a pound of weight is wasted ; the outward thrust of the 
 vault-arches within is so nicely calculated that the mass of 
 the buttress is just sufficient to ensure the stability of the 
 building ; its form also, thin and deep, shows the con- 
 summate skill with which the fifteenth-century builders 
 distributed the abutment of their arches. The very 
 pinnacles that strike the eye as mere ornaments contribute 
 their part to the stability of the building by weighting the 
 buttress against the outward thrusts of the arches within. 
 
 The architecture of the century following the Reforma- 
 tion is usually described as debased Gothic, and is regarded 
 as the final stage of a gradual process of degeneration that 
 began early in the fifteenth century ; from this point 
 of view the revival of the Classic Orders was an attempt 
 to replace a decadent style by a new and vigorous system. 
 
 That there were signs of degeneracy in fifteenth- 
 

 FIG. 17. THE DIVINITY SCHOOL (t. 1450) 
 
68 THE HISTORY OF ARCHITECTURE 
 
 century architecture cannot be denied ; a failing sense 
 of beauty is seen in the carved foliage, a lack of restraint 
 and lower ideals in the profusion of easily executed details, 
 and a stinting of design in the repetition of similar forms. 
 And it is significant that the ornamental details of the 
 Classic Orders were revived long before their structural 
 principles, as if the builders sought to retain Gothic 
 construction while abandoning its debased ornament. 
 
 In spite of this a good deal might be said to show that 
 sixteenth-century Gothic was not a debased style, that the 
 revival of Classic architecture was the cause and not the 
 result of its downfall, and that but for that revival our 
 cities to-day would be as beautiful as in the Middle Ages. 
 Even in Tudor times some of our finest examples of 
 Gothic were still being designed ; the present nave and 
 choir of S. Mary's and the bell- tower of Magdalen 
 the finest tower of the Middle Ages were built in the 
 last years of the fifteenth century, while Christ Church 
 Hall is more than a quarter of a century later. These are 
 no more debased architecture than is New College. And 
 that even the seventeenth century could produce Gothic 
 work worthy to set beside these may be seen in Christ 
 Church stairway and Wadham Chapel. 
 
 It may be said that these are isolated examples and that 
 the vast majority of Elizabethan and Jacobean buildings 
 have the lintelled windows and depressed door-arches 
 that are understood to be the marks of debased Gothic. 
 It is true that the Fellows' Buildings at Merton have 
 not the traceried windows, the shafted doorways and ex- 
 quisite carved foliage of the Chapel; but this is no evidence 
 that they are a debasement of the earlier style. They 
 are rather to be regarded as an attempt to modify the 
 details of mediaeval architecture to meet the building 
 
70 THE HISTORY OF ARCHITECTURE 
 
 requirements of a later age. With, the new notions of the 
 Renaissance came new needs ; the simple plan of the 
 church nave that had served for the early manor-house 
 could serve no longer ; far more complicated planning was 
 now necessary. With the Reformation, too, art ceased to 
 be exclusively associated with religion ; no more churches 
 were needed, and monastic building came to an end. 
 It was inevitable that, with the development of civiliza- 
 tion, ecclesiastical and domestic buildings should diverge 
 more and more from a common type ; but it is surely 
 wrong to brand the newer variation as a debased form. 
 
 The abandonment of the pointed arch is generally 
 regarded as the proof and sign of degeneration. This 
 idea is due to the mistaken belief that the pointed arch 
 is the fundamental characteristic of Gothic architecture. 
 A vaulted space ought always to be lighted by windows 
 corresponding in form to the arches of the vault, but in 
 unvaulted buildings this necessity does not exist, and 
 so square-headed openings are not uncommon even in 
 the finest period of the style. In the small windows of 
 domestic buildings they are the most obvious and con- 
 venient form. That they need not necessarily be merely 
 utilitarian every Jacobean manor-house proves. It is when 
 they are inserted in the walls of ancient churches that they 
 appear debased. But after the shock of the Reformation 
 had passed, the builders appreciated this, and in the 
 chapels of Wadham, Oriel, University, and Lincoln, as 
 well as in those of Balliol and Exeter, which have since 
 been rebuilt, they attempted to revive the tracery of the 
 mediaeval window. That the results were less beautiful 
 than mediaeval work was an inevitable effect of the 
 Reformation. 
 
 The tower of Yarnton Church (1611) and the Chapel 
 
72 THE HISTORY OF ARCHITECTURE 
 
 of Water Eaton (1600), if compared with the adjoining 
 manor-houses, will also serve to illustrate the seventeenth- 
 century attempt to adapt Gothic architecture alike to 
 domestic and ecclesiastical needs. 
 
 When, in the Middle Ages, the church window was 
 the house window, the men who lived in houses worthy of 
 the name were no more numerous than the churches. 
 A more democratic age could not build church windows 
 in all its dwellings, and would not continue to build them 
 in a few. It therefore evolved a modified form for domestic 
 buildings, which should not challenge comparison with 
 the house of God, which was adapted both to manor- 
 house and cottage, and was convenient and beautiful 
 in either. In spite of all that has been written about 
 debased Gothic, I think that the last phase was a natural 
 development and not a debasement of the style. The 
 Renaissance of Classic literature brought into contempt 
 the building styles of the Middle Ages no less than the 
 writings of the Schoolmen, even though they had not 
 merited it ; the downfall of Gothic architecture was 
 brought about, not by the slow process of degeneration, 
 but by a blind enthusiasm for everything Classic. Now, 
 after three centuries of arrogant and ugly buildings, we 
 are beginning again where the Elizabethans left off. 
 
 CHAPTER V 
 
 THE RENAISSANCE AND AFTER 
 
 IN the sixteenth century, Rome was the Mecca of 
 scholars. It was natural that the interest in Classical 
 literature should extend to architecture, and, perhaps, 
 equally natural that the mediaeval systems should fall into 
 
74 THE HISTORY OF ARCHITECTURE 
 
 contempt. In Italy itself, where Gothic had always been 
 an exotic, the style was early abandoned, and the writings 
 of Vitruvius became the gospel of a new school named 
 the Palladian, after Palladio, its leading apostle. 
 
 This Italian school was afterwards to provide models 
 for the western countries ; but at first the nations by 
 whom Gothic was invented, seemed loath to abandon it 
 completely ; perhaps it was the architects and not the 
 wish that was lacking. Native craftsmen built the Chapel 
 of Henry VII at Westminster, but an Italian architect 
 was employed to design his tomb ; it was not until a 
 supply of English architects, trained abroad, was available 
 that large buildings were designed in the new style ; 
 until the middle of the seventeenth century buildings 
 were still Gothic, except that a doorway, a chimney-piece, 
 or a porch, might be added by an Italian craftsman. 
 Among our earliest examples of the mixture of styles 
 are Anthony Forster's tomb at Cumnor, and Bishop 
 Jewel's porch at Sunningwell. By the end of the sixteenth 
 century there were Classic details in every building, but 
 where it was purely the work of native craftsmen they were 
 not sufficient to un-Gothicize the effect. Thus Wadham 
 College, built by Somersetshire masons, has a far more 
 Gothic appearance than the back quadrangle of S. John's, 
 which was designed by Inigo Jones only a few years later. 
 
 Inigo Jones (1573-1652) was the first great English 
 architect to break definitely with the older traditions. 
 He had studied abroad the architecture of the Italian 
 Renaissance, and on his return to England was com- 
 missioned by James I to build his new palace of White 
 Hall. His design, which was never completed, has no 
 Gothic features ; it is purely Classic, like his Gateway 
 to the Botanical Gardens (1633). His work marks the 
 
76 THE HISTORY OF ARCHITECTURE 
 
 final break with the system of the Middle Ages, not alone 
 with the building style, but with the conditions under 
 which the work was carried out. Wykeha*m or Merton 
 had been content to lay down the general plan of his 
 buildings, leaving the form of minor details to the 
 individual workman the grotesques in the cornice of 
 Merton Chapel, for example, were obviously designed by 
 the man that cut them and not by Merton or his master 
 builder ; but the new school of architects, like the 
 ancients they copied, worked out the complete design 
 on paper, down to the minutest details, and left the 
 workman no responsibility but that of accurately copying 
 them. To Inigo Jones succeeded Sir Christopher Wren, 
 of whom it might be said in Broad Street as truly as 
 in S. Paul's, si monumentum requiris circums-pice. The 
 Chapel of Brasenose College (1656) is sometimes attributed 
 to Wren, who was a fellow of All Souls at that date. But 
 it is hardly probable that an architect with his knowledge 
 of principles would have designed such a mixture of 
 Classic and Gothic details. It was his work in Oxford 
 that put a definite end to the lingering Gothic and 
 inaugurated a period of unmixed Classic architecture. 
 Unmixed, that is, with any Gothic features ; but the 
 buildings of Wren, like those of Palladio, are a return to 
 that anomalous system which prevailed when the Roman 
 architects of the first century were seeking to combine 
 the forms of the Classic Orders with the principles of 
 arcuated construction. So in the front of the Sheldonian 
 Theatre (1666) Wren supports a Corinthian entablature 
 by means of arches, and disguises their abutment in the form 
 of columns that appear to be carrying the weight. His 
 finest work in Oxford is the Chapel of Queen's College, 
 especially interesting for its revival of the Roman apse 
 
FIG. 22. ALL SAINTS' CHURCH (c. 1709) 
 
78 THE HISTORY OF ARCHITECTURE 
 
 in its chancel. Wren was also the first English architect 
 to revive the Roman dome. His earliest experiment, the 
 little dome of the Sheldonian Theatre, 1 is the small 
 sister of his mighty one at S. Paul's, and the beautiful 
 one of Christ Church. 
 
 But it is not merely by his own buildings that Wren left 
 his mark on Oxford ; his influence is seen in the work 
 of other architects, in the front quadrangle of Queen's 
 College and the back quadrangle of All Souls, built by 
 his pupil Hawksmoor, in the Chapel of Trinity, designed 
 by Dean Aldrich, and in the great dome of the Radcliffe, 
 built by Gibbs in 1750. Dean Aldrich of Christ Church 
 represents a new type of architect, the amateur, made 
 possible by the new condition that the directing mind 
 need not be that of a craftsman. The most famous of 
 the amateurs was Sir John Vanbrugh 
 
 Lie heavy on him earth, for he. 
 Laid many a heavy load on thee. 
 
 Perhaps the heaviest, in proportion to its area, is the 
 Clarendon Building, built in 1709 ; but Blenheim Palace 
 is his best known work and perhaps his ugliest. Dean 
 Aldrich designed Peckwater Quad at Christ Church, and 
 is credited with the design of All Saints' Church, 1710. 
 The great interest of the building is in its spire, an 
 essentially Gothic feature grafted on to a Classic base. 
 The spire, which was the one Gothic structure built 
 solely for display, was naturally the one retained by the 
 architects whose models were the costly and magnificent 
 buildings of imperial Rome. Wren's London spires are, 
 of course, known to every one ; few are so near to the 
 mediaeval form as the spire of All Saints'. It is as if Gothic 
 1 Much modified in 1838. 
 
THE RENAISSANCE AND AFTER 79 
 
 struggled in its grave. The principle of verticality, indeed, 
 never ceased to struggle in Oxford ; it sprang up again 
 in Hawksmoor's towers at All Souls perhaps the influ- 
 ence of the old Gothic surroundings was too strong to 
 be resisted ; certainly that influence must have been felt 
 as a disturbing force by eighteenth-century architects in 
 Oxford. When the New Buildings were added to 
 Magdalen in 1733, it was proposed to pull down all the 
 Gothic work and rebuild the college in harmony with 
 the new block. Fortunately lack of funds caused the 
 design to remain upon paper. 
 
 So for two centuries and a half the Classic style pre- 
 vailed in Oxford, and, of course, throughout Western 
 Europe (the garden quadrangles of New College (1684) 
 and Trinity (1665), suggested by the plan of the new 
 palace of Versailles, serve to remind us that the Classic 
 revival was common to civilized Europe). The ancient 
 Orders had been welcomed with enthusiasm, but from the 
 first it might have been foretold that they could not 
 satisfy the architectural needs of a modern nation. The 
 true Renaissance of architecture was in the twelfth 
 century, when the western peoples took the arcuated 
 system of Rome and developed from it the glorious 
 Gothic style ; the sixteenth-century architectural 
 Renaissance was a false one ; the real Renaissance of 
 that date was literary ; and the birth of literature and 
 architecture can never coincide in time ; architecture 
 is the earliest, literature the latest, of the arts. 
 
 Classic architecture became the symbol of plutocracy. 
 It was essentially a style for the great and the rich ; it 
 could not condescend to the resources of the local quarries; 
 its massive lintels and its columns must be fetched from 
 afar ; its associations demanded scholarship in the 
 
8o THE HISTORY OF ARCHITECTURE 
 
 architect, and its refinements and symmetry demanded 
 exceptional skill in the craftsman ; moreover, the price 
 to be paid for it was the utter subordination of the indi- 
 vidual workman. When its conditions cannot be granted, 
 when its price cannot be paid, it becomes a mockery 
 like the dwarf columns in the fronts of houses near 
 Queen's College. 
 
 Before the end of the century it had become a pompous 
 absurdity, in the dull, heavy inanity of Worcester College, 
 for example, or in Wyatt's Gateway to Canterbury Quad. 
 Early in the nineteenth century its imbecility became 
 too obvious to be longer tolerated ; and it was ' put 
 away '. 
 
 CHAPTER VI 
 
 THE GOTHIC REVIVAL 
 
 THE Classic style had been tried and found wanting. 
 It had no affinity with the ideas of a modern people, 
 and it was not adaptable to modern needs ; it forced 
 the architect to sacrifice comfort and convenience for the 
 sake of appearance ; as Pope said of Blenheim Palace : 
 
 'Tis mighty fine ; 
 
 But where d'ye sleep and where d'ye dine ? 
 I find from all you have been telling 
 That 'tis a house but not a dwelling. 
 
 Disciples of the Oxford Movement denounced the 
 style as Pagan, and preached a return to the architecture 
 of the ages of faith. It certainly was an essentially un- 
 Christian style, not because it was the style of the heathen 
 temples, but because it enslaved the workman, and was 
 
THE GOTHIC REVIVAL 
 
 81 
 
 purse-proud and arrogant. But the Gothic revivalists 
 fell into the error of the Renaissance architects ; led by 
 blind enthusiasm, they began to reproduce Gothic 
 buildings as if the architecture of the thirteenth century 
 
 FIG. 23. SEVENTEENTH-CENTURY HOUSE, BALLIOL COLLEGE 
 
 could satisfy the needs of the nineteenth. The originators 
 of the great building styles, of Egypt, of Greece, of Rome, 
 and of Mediaeval Europe, had evolved them, not by 
 copying, but by taking the principles of older work and 
 applying them to their own newer needs and circum- 
 stances. In no other way can a worthy style of architecture 
 be produced. 
 
 785 F 
 
82 THE HISTORY OF ARCHITECTURE 
 
 The nineteenth century, with its enormous increase 
 in wealth and population, with its Reform Bills and Educa- 
 tion Acts, saw conditions utterly different from any 
 previously existing. What it urgently required was a 
 system of domestic architecture suited to the physical 
 and spiritual needs of a modern democratic nation. Such 
 a system, of course, had never been evolved, and could 
 not be copied. 
 
 In the Middle Ages the vast majority of domestic 
 buildings had been made of timber a framework of 
 posts and beams with the interstices filled in with wattle 
 and daub, or with laths covered with plastered clay. 
 In the villages near Oxford many a ruined cottage 
 can be seen with the laths or wattle exposed where 
 the clay has broken away ; and many more timber- 
 framed houses are still inhabited, but bricks have been 
 substituted for the original plaster. Bishop King's 
 Palace and the old house behind the west front of 
 Balliol are fine examples of the timber-framed houses 
 of Elizabeth. 
 
 With the rise of a yeoman class, with the growth of 
 wealth and population, with the dissemination of learn- 
 ing, there grew up in later Tudor times a general desire 
 for more substantial and comfortable dwellings. It ill 
 becomes the present utilitarian age to condemn Jacobean 
 builders for admitting that houses could no longer be 
 built in the style of churches, and for modifying Gothic 
 forms to suit purely material needs ; it was still more 
 absurd for the Early Victorians to brand their work as 
 debased. Nevertheless, the cry was for Gothic, the whole 
 Gothic, and nothing but Gothic, in church, chapel, gaol, 
 county court, school, and city dwelling. Practically the 
 only models were the mediaeval churches, and the architects 
 
84 THE HISTORY OF ARCHITECTURE 
 
 * tried to answer the demand by putting church doorways 
 and windows into all their buildings, as if pointed arches 
 to the openings could make a building Gothic. In churches 
 the mediaeval style might seem as suitable now as in the 
 past ; like needs produce like results ; faiths and dogmas 
 change slowly, and the members of Exeter College may 
 reasonably worship in the twentieth century in a replica of 
 a chapel built in Paris in the thirteenth. Yet even to copy 
 a Gothic church is to ignore the fact of the Reformation ; 
 a mediaeval church was not planned for congregational 
 worship it was a congeries of chapels and chantries, 
 each with its own altar, each divided from the others. 
 A modern church should be a single great auditorium ; 
 it is futile to attempt to revive the spirit of the Middle 
 Ages (even if it were desirable) by reverting to its type 
 of church-building. The numerous piers of an ancient 
 church exist, not because the builders desired to obstruct 
 sound and sight, but because they could not roof the 
 building without them. Modern architects have no 
 such excuse ; in using them they are but copying the 
 weakness of a more primitive style. S. Paul's Church, 
 ugly as it is, is a more intelligent attempt to meet 
 modern needs than any of the Gothic imitations of North 
 Oxford. 
 
 Still, the mediaeval revivalists had their way. In 1826 
 S. Clement's Church was rebuilt on a new site in a style 
 contemporary with the First Crusade ; twenty years 
 later the County Gaol was built in the same style. A few 
 people understood that architecture was still on the 
 wrong road, and in 1846 Cockerell built the Taylorian 
 Buildings in the pure Ionic style, apparently in the belief 
 that the failure of the Renaissance work had been due 
 to the imitation of Roman corruptions. Pure or corrupt, 
 
86 THE HISTORY OF ARCHITECTURE 
 
 the Classic style had become impossible, and the great 
 epidemic of Gothic broke out everywhere. 
 
 Sir Gilbert Scott was the master craftsman of the 
 revival, and Ruskin was its chief apostle. Scott's first 
 work was the Martyrs' Memorial, copied from the 
 Eleanor Crosses ; at the same time, 1846, he rebuilt the 
 north aisle of S. Mary Magdalene Church in the style of 
 the early fourteenth century. Then (1856) he blew up 
 with gunpowder so. well preserved was the building 
 that nothing less would serve the Chapel of Exeter 
 College, and replaced it by the present reproduction of 
 the Sainte Chapelle. He also added the new wing on 
 Broad Street, which remained for many years the most 
 imbecile building in Oxford, until it was eclipsed by his 
 Holywell front of New College. His contemporary, Sir 
 Charles Barry, the architect of the Houses of Parliament, 
 added a west wing to University College in 1843, in the 
 style of four centuries earlier, while Mr. Butterfield, in 
 1857, rebuilt the front of Balliol in imitation of Edward Fs 
 castle at Carnarvon, and in 1864 cut down the Grove 
 of Merton to build a new thirteenth-century block. He 
 also built Keble College in 1870, using a material that was 
 unknown to the thirteenth-century builders whose work 
 he imitated. 
 
 Meanwhile, Ruskin had discovered that Gothic architec- 
 ture had reached its highest perfection in mediaeval 
 Venice, and, inspired by him, Mr. Waterhouse built the 
 University Museum and the Meadow Buildings of 
 Christ Church in imitation of Venetian Gothic. 
 
 Other architects had their pet periods to reproduce ; 
 Mr. Buckler, in 1854, had built Magdalen College School, 
 and in 1856 rebuilt the front of Jesus College in the 
 Perpendicular style of the fifteenth century ; the Church 
 
88 THE HISTORY OF ARCHITECTURE 
 
 of S. Peter le Bailey was rebuilt on a new site in a style 
 contemporary with Edward III ; and that of Holy 
 Trinity in the Lancet style of the early thirteenth century. 
 The architect of S. Barnabas Church, Sir A. Blomfield 
 (1869), more logical than the Gothicists, decided that if 
 ancient types were to be reproduced, the most suitable 
 was the basilican church of the early Christians, which 
 he therefore proceeded to copy. Later in the century 
 Mr. Bodley added S. Swithin's Buildings to Magdalen, 
 making the first attempt at applying Gothic principles 
 to modern conditions instead of merely copying Gothic 
 details ; these buildings are therefore among the most 
 successful in modern Oxford. 
 
 But the architect whose name will be known to posterity 
 as the representative of the builders of our day is Mr. T. G. 
 Jackson. Mr. Bodley seems to have been the first to 
 realize that a domestic type of architecture was the chief 
 need of his day, and since the only domestic type of Gothic 
 was the Jacobean manor-house, he reverted to that in 
 building (1879) tne Master's Lodge at University College, 
 ' perhaps the most beautiful modern house in Oxford '. 
 Mr. Jackson followed his lead, and has sought inspiration 
 from the same source. His great work is the new Examina- 
 tion Schools ; other examples of buildings in the same 
 style are his quadrangle at Trinity College with the 
 President's House (1887), his new buildings at Hertford 
 and Merton, and the Library of the University Museum ; 
 the Town Hall by Mr. Hare, and the Indian Institute 
 by Mr. Champneys. 
 
 The church builders still continue to imitate the 
 mediaeval styles or rather to reproduce their details. 
 The poverty of churchmen compels them to build in 
 brick, but instead of attempting to discover how to evolve 
 
THE GOTHIC REVIVAL 89 
 
 a really architectural style of using that material, they 
 face their buildings with a skin of stone, which looks just 
 as well and costs half as much as if stone had been employed 
 throughout in the ancient manner. Thus we have lately 
 seen the building of two ' Norman ' churches in brick, 
 S. Andrew's and the new Roman Catholic Church, one 
 disguised with rubble the other with flint. In order to 
 give greater verisimilitude S. Andrew's has been provided 
 with a sham vault of plaster. 
 
 Meanwhile the original problem of modern architecture 
 remains unsolved : what style of domestic building has 
 been evolved to meet the needs of a great, educated, 
 democratic people ? The houses of East Oxford supply 
 an answer. What will posterity think of it, what deduc- 
 tions draw from it ? I have in mind a row of ten houses 
 on a main road. They form a block of forty pigeon-holes 
 under a single roof ; four pigeon-holes accommodate a 
 family ; the oblong front of the block is pierced by forty 
 oblong openings for doors and windows : there is the 
 complete plan and elevation of dozens of house-blocks 
 representing no more design than is required to build 
 a rabbit-hutch. 
 
 To be happy in a dog-kennel one must be either a dog 
 or Diogenes ; and Diogenes was already a philosopher 
 when he took up his residence in a barrel, or he would 
 never have become one. Children brought up in mean 
 streets of dull house-fronts have but a poor chance of 
 developing that love for the beautiful which, more than 
 any other attribute, distinguishes men from beasts. 
 
 CONCLUSION 
 
 Architecture has been too long the plaything of the 
 antiquarian. It is time that it was recognized as a matter 
 
9 o THE HISTORY OF ARCHITECTURE 
 
 of vital concern for every citizen ; mere building is so 
 recognized, and every house, nay, every piggery, has 
 to be constructed in accordance with the conditions 
 necessary for man's physical wellbeing. But what 
 building by-laws recognize the existence of a spiritual 
 instinct in man that craves for beauty and dignity in 
 human surroundings ? Yet it is by virtue of that instinct 
 that man is man ; if it is starved or stultified he descends 
 again to the level of the brutes from which he sprang. 
 Where there is no vision the people perish. 
 
 Only a great revival of art can save the modern nations 
 from degenerating through specialization and materialism 
 into communities of ants, and that revival can best begin 
 in architecture, the first, the most necessary, and the 
 most universal of the arts. If the average man had only 
 as much knowledge of the principles of architecture as 
 he has of sanitation, if he only realized that ugliness is 
 as pernicious to the mind as stinks to the body, our cities 
 might soon become as beautiful as they are healthy ; an 
 ugly building would no more be tolerated than an 
 insanitary one. 
 
 Every building raised becomes a part of the public 
 environment ; and though its owner may make it a 
 pleasant dwelling for himself, he owes it to society to 
 make it also a pleasant object to his neighbours. That is 
 the price he must pay for his alteration in the landscape. 
 He has no more right to offend the public eye than he 
 has to endanger the public health. 
 
 ' You cannot make men artistic by Act of Parliament : ' 
 ' Man is born of woman, not of the Local Government 
 Board.' As usual, truth is sacrificed to an epigram. Man 
 is born of woman, but he is developed by his environment, 
 over which society can, and to some extent does, exert 
 
THE GOTHIC REVIVAL 91 
 
 control. Are we, who have produced more great poets 
 than any other nation, so oblivious of their teaching that 
 we have no realization of the educational importance of 
 beautiful things ? Architecture, especially in towns, is a very 
 dominant part of environment ; if it is honest, beautiful, 
 and dignified, it must have a like effect upon the minds 
 of those brought up amongst it ; if it reflects selfishness, 
 cheap ostentation, and bad taste, it must leave like im- 
 pressions on plastic minds. The indifference of the public 
 to such a matter of universal concern is as lamentable 
 as it is astonishing. 
 
 There are, it is true, some signs that the sun of art 
 which went down into the black night of materialism 
 a hundred years ago may shine again upon a later genera- 
 tion : we have garden suburbs, we are learning to preserve 
 our ancient buildings, and we are asking for picturesque- 
 ness in our suburban villas. But we lack knowledge of 
 architectural principles to criticize our architects, to 
 encourage the good and send the bad out of business. 
 In North Oxford there are hundreds of modern houses 
 fondly believed by their occupiers to be * half-timbered ' 
 like the timber-framed houses of the sixteenth century. 
 Really they are ordinary brick houses with boards nailed 
 to their fronts, having only the sham picturesqueness of 
 stage-scenery. Iron now enters largely into the building 
 of our great shops, but instead of admitting the fact and 
 devising an architectural use for the new material, the 
 builder disguises construction and hides his girders behind 
 wooden shop-fronts. We still allow any individual who 
 has the money to do so to ruin a beautiful view by raising 
 a high blank wall around his grounds, as has happened 
 lately on Boar's Hill. 
 
 Finally, and worst of all, our factories are as hideously 
 
92 THE HISTORY OF ARCHITECTURE 
 
 utilitarian as ever. When workmen cease to be ' hands ', 
 they will strike for due regard for their spiritual no less 
 than for their material needs. 
 
 Because a man has a shop to mind 
 
 In time and place, since man must live, 
 
 Need spirit lack all life behind, 
 All stray thoughts, fancies fugitive, 
 All joy except what shop can give ? 
 
 I want to know a butcher paints, 
 A baker rhymes for his pursuit, 
 
 Candlestick-maker much acquaints 
 His soul with song, or haply mute, 
 Blows out his brains upon the flute. 
 
 But shop each day and all day long 
 Friend, your good angel slept, your star 
 
 Suffered eclipse, Fate did you wrong ; 
 From where these kinds of treasures are 
 There should our hearts be Christ, how far ! 
 
PART II 
 THE GRAMMAR OF ARCHITECTURE 
 
 CHAPTER I 
 
 INTRODUCTORY 
 
 WE have more than once suggested that the parts of 
 a building are to be studied in relation to the roof, since 
 all subserve the essential function of providing a sheltered 
 enclosure. It is proposed in this section to make an analysis 
 of English building construction, studying the details of 
 its anatomy in their relations to each other and to their 
 common purpose, tracing the origin of its parts and 
 their modifications throughout the historical period. 
 The changes in detail which mark the different periods 
 are not to be regarded as mere changes of fashion due to 
 changing aesthetic taste ; they are almost always traceable 
 to attempts to improve the practical utility of the 
 building to provide a more stable roof, to improve the 
 lighting, to remove impediments to sound and light, or 
 to economize material and simplify construction. The 
 modern student of architecture, like the post-Darwinian 
 student of nature, must study the details of structure in 
 relation to function ; and, like him, he will find far more 
 delight in the intellectual exercise than in the mere 
 memorizing of isolated phenomena. 
 
 The church was the typical building of the Middle 
 Ages ; castles and manor houses might be built, but the 
 church tower was the model for the one and its nave 
 for the other. From the sixth century to the sixteenth 
 the aims and energies of architecture were directed to the 
 
94 THE GRAMMAR OF ARCHITECTURE 
 
 improvement of God's House, the men who laboured 
 upon it living in huts of straw and clay. The study of 
 church anatomy is, therefore, the study of mediaeval 
 building construction. 
 
 We shall have to consider in order the plan of the church, 
 with its main divisions and their origin, the roof, both the 
 stone vault and its timber protection, the buttresses upon 
 which its thrusts are concentrated, the arches and the 
 piers that carry them, the walls with their openings for 
 doors and windows, and finally, the details of mediaeval 
 ornament. In each instance we shall trace the subsequent 
 history of the form through the Renaissance period down 
 to our own day. 
 
 THE CHURCH PLAN 
 
 The origin of the plan of the Christian church is to be 
 found in that of the Roman Basilica or Court of Justice. 
 This was an oblong hall with a semicircular recess, the 
 apse, at one of its ends, in which sat the judge and his 
 attendant officials. Basilica were of two types, represented 
 in Oxford by the churches of S. Paul and S. Barnabas. 
 In the simpler and smaller examples a single roof spanned 
 the building from wall to wall ; but where greater width 
 was required for the accommodation of large audiences it 
 was necessary to divide the hall longitudinally into three 
 parts, by means of piers ; arches were built upon these 
 to carry a wall for the support of a medial roof, and the 
 side divisions, or aisles, were roofed separately, either by 
 lean-to roofs of timber or by cross-vaults. 1 The apse was 
 covered by a half-dome resting upon its semicircular wall. 
 
 1 The great arched entry to the Clarendon Press will serve to 
 illustrate this arrangement. The central roadway is roofed with 
 a barrel vault, flanked by groined vaults over the side passages. 
 
THE CHURCH PLAN 95 
 
 Both types were copied in the churches of the early 
 Christians, the hall becoming a nave for the worshippers, 
 and the apse a sanctuary for the officiating priests ; it 
 was cut off from the body of the church by screens 
 (concetti), and, hence, came to be known as the chancel ; 
 the foundations of a small Romano-British church on 
 the unaisled basilican plan were recently unearthed at 
 Silchester ; in the more important churches of Italy the 
 aisled plan was general. 
 
 The early Romanesque builders, Saxon and Norman, 
 built their smaller churches upon the simpler plan, 
 but the larger churches were aisled, and aisles as well as 
 nave ended in apses at the east. The foundations of three 
 apses in the Canons' Gardens of Christ Church seem 
 to indicate that the original church of S. Frideswide's 
 nunnery was built on this plan. 
 
 But the difficulty of building semicircular walls, and 
 still more of roofing the apse with a semi-dome, led in 
 many cases to the building of the square-ended chancels 
 that had been the rule in the primitive timber-built 
 churches. So we have the plan of Elsfield Church (c. 1220), 
 which was also the original plan of S. Peter's (c. 1120). 
 This arrangement, at first adopted for convenience in 
 building the smaller churches, became the prevailing 
 fashion in England before the end of the twelfth century. 1 
 The east end of our own Cathedral is a conjectural 
 reconstruction of the original plan of 1160. 
 
 The splendid Norman mind was not content with 
 such a simple building as the basilica for the churches of 
 bishops and abbots ; the demands of an increasingly 
 
 1 On the Continent the apse was retained in Gothic work 
 (Fig. 25). At Tidmarsh, near Pangbourne, there is a rare English 
 example of the thirteenth century. 
 
FIG. 27. CENTRAL SPIRE OF CATHEDRAL (c. 1250) 
 
THE CHURCH PLAN 97 
 
 elaborate ritual, too, and the provision of altars for saints, 
 made necessary a more complex organization of the parts 
 of the building. To meet the new ne.eds the cruciform 
 plan was evolved by carrying a second nave across the 
 first at its junction with the chancel ; this transept was 
 usually unaisled, but in our Cathedral we have the idea 
 carried to its full development in an aisled transept 
 crossing an aisled nave and chancel. Over the square 
 of intersection a tower was built, not only to give dignity 
 to the building, and effective grouping, but to annihilate 
 by its weight the thrusts of the arches collected at that 
 point ; and even in churches where there were no arcades 
 the four walls of the central tower greatly simplified the 
 difficulties arising from the meeting of the four roofs, 
 of nave, choir, and transepts ; each could be made to 
 end in a gable against one of the four walls of the tower. 
 
 The idea of a central tower was dear to the Romanesque 
 builders, and often, as at Iffley, we find one where there 
 is no transept. In the great cruciform churches a central 
 tower was always structurally necessary, and even in 
 village churches, as at Beckley, the Gothic builders would 
 occasionally indulge in the luxury. 
 
 But in the simpler churches, after the twelfth century, 
 the tower was usually built at the west end, where it was 
 of greater structural value ; its weight received the thrusts 
 of the nave arcades, which were resisted at the east by 
 the walls of the chancel. The plan of the typical Gothic 
 church is thus that of the aisled basilica, with a western 
 tower, and a square chancel instead of an apse. This plan, 
 however, is commonly the result of additions to an origin- 
 ally simpler building. S. Peter's Church, for instance, 
 as first built in 1120, had nave and chancel only. A 
 hundred years later it was enlarged by the addition of 
 
 785 G 
 
98 THE GRAMMAR OF ARCHITECTURE 
 
 aisles to the north of both ; this proved sufficient for the 
 needs of the parish, and so no south aisle was added. 1 
 At Holywell, on the other hand, a south aisle was built 
 to the Norman nave in the thirteenth century, and another 
 to the north in the fifteenth, while at S. Giles's, not only 
 were aisles added to the nave in the thirteenth century, 
 but a choir-aisle was built to the south of the chancel. 
 This choir-aisle probably served the purpose of a Lady 
 Chapel ; the thirteenth century had a great enthusiasm 
 for the worship of the Mother of Christ, and many chapels 
 were then added to chancels in her honour. In minster 
 churches the Lady Chapel was usually built on to the 
 east end; but the church of S. Frideswide was so near 
 the city wall that there was no room to extend it eastward, 
 and the Lady Chapel was therefore built to the north 
 of the choir-aisle. So, too, when the Latin Chapel was 
 contemplated in the fourteenth century it was necessary 
 to build it still further north ; for the monastery occupied 
 the ground to the south. 
 
 The plan of every mediaeval church reflects the worship 
 of saints, which was an integral part of the religion of the 
 age. In Romanesque churches the bones or other relics 
 of holy men were usually preserved in a crypt built 
 beneath the floor of the chancel. This ' hidden place ' 
 was intended to reproduce the catacombs in which the 
 early Christians had been buried. At certain festivals 
 the relicswere displayed, and the worshippers were allowed 
 to walk round the crypt and to look upon them as they 
 passed. In the crypt of S. Peter's Church are traces of 
 two stairways on either side of the chancel arch ; by one 
 of these the faithful entered, and, passing round the outer 
 
 1 But a small transeptal chapel and a porch were added in the 
 fifteenth century. 
 
THE CHURCH PLAN 99 
 
 aisle, returned by the other. After the twelfth century, 
 relics were generally transferred to shrines in chapels built 
 for them in the church above. Crypts were still built 
 there are examples beneath the churches of S. Aldate and 
 S. Mary Magdalene and the Chapel of All Souls but they 
 had no ceremonial use ; they were often charnel houses 
 where bones dug up in the churchyard were preserved. 
 
 The mediaeval builders had planned so generously 
 that few churches were necessary till long after the 
 Reformation. The seventeenth and eighteenth centuries 
 were the age of chapel-building by the dissenting sects. 
 Built, unlike the churches, only for congregational 
 worship, the chapel was rightly planned as a great rect- 
 angular hall ; All Saints' Church, 1710, is typical of the 
 post-Reformation House of Worship ; the resources of 
 the builders enabled them to dispense with obstructive 
 piers and to build the roof of a single span. The later 
 Gothic builders had achieved the same feat in the 
 Divinity School, and in King's College Chapel at Cam- 
 bridge. Modern architects, in returning to the aisled plan, 
 are putting back the clock. The eighteenth-century 
 churches and chapels are certainly ugly, but it is not 
 because they are badly planned. 
 
 THE COLLEGE PLAN 
 
 The origin of the quadrangular arrangement of col- 
 legiate buildings is by no means certain. Wykeham's 
 quadrangle at New College set the type for all later 
 colleges, and his plan is commonly supposed to have been 
 derived from that of the mediaeval monastery. Some 
 idea of the monastic plan may be gathered from the 
 cloister quadrangle of Christ Church. 
 
 Seven times a day to praise God was the principal end 
 
ioo THE GRAMMAR OF ARCHITECTURE 
 
 of a monk's existence ; a monastery was, therefore, 
 primarily a great church with adjacent buildings in which 
 those who served it might rest and eat, and perform 
 necessary tasks in the intervals of devotion. These 
 subsidiary buildings were usually placed on the south side 
 of the church, and were thus protected by its lofty ridge 
 against the cold north winds. They were arranged in 
 three blocks, all facing inwards upon a central cloister 
 garth and turning dead walls to the outside world. We 
 have seen that this was the plan of the Romano-British 
 villa ; it was also the plan of the fortified manor house ; 
 and it was obviously the best possible arrangement for 
 shutting out enemies, whether temporal or spiritual. 
 
 On the side of the quadrangle, remote from the church, 
 were the commissarial buildings, the kitchen, buttery, 
 and refectory ; the refectory of S. Frideswide became 
 the library of Christ Church, and was converted into rooms 
 in 1775, after the building of the new library in Peck- 
 water. Opening out of the south transept of the church 
 was the sacristry, and next was the Chapter House, in 
 which the monks met daily to discuss the affairs of the 
 monastery, to punish offences against discipline, or to 
 receive orders from the abbot or prior. An arched passage 
 or slype communicated with the graveyard outside the 
 cloister, and near it was the mortuary. On this side, too, 
 was the monk's day-room, and above the whole block was 
 their dormitory, opening at its northern end into the 
 transept of the church, into which they descended by 
 steps for the midnight offices. On the remaining side was 
 the lay-brothers' day-room with their dormitory above ; 
 no trace of this block remains at Christ Church. 
 
 All round the inner quadrangle ran a covered way, 
 usually vaulted in stone, and known as the cloisters ; 
 
t r 
 
 CHUR 
 
 }. . I . 
 . 
 
 CLOISTER CAR' 
 
 I L.- 
 
 BUTTtRY I I * * 
 
 ' | I KtTCMrM 
 
 i*4| t*v 
 
 * 
 
 P REFECTORY (i 
 
 1 * I 
 
 
 
 I 
 
 FIG. 28. THE MONASTIC PLAN 
 
102 THE GRAMMAR OF ARCHITECTURE 
 
 it was open to the central garth, and in each of its arched 
 bays was fitted a ' carrel ', a small wind-shelter, serving 
 as a study in which a single monk could read or write. 
 
 These were the only essential monastic buildings ; 
 outside, there were doubtless barns and stores, an in- 
 firmary, and usually a mill ; as wealth increased and 
 devotion decreased, the abbot and prior built separate 
 houses for themselves, and a great gateway, with porter's 
 lodge, and guest-house was added ; but, originally, cloister 
 and monastery were synonymous. 
 
 Now the college plan bears only a superficial resemblance 
 to that of a monastery ; and that resemblance is probably 
 due to like needs producing like results rather than to 
 conscious imitation. In an age when emeutes were 
 frequent in every city, studious quietness could only be 
 obtained by the adoption of the ancient plan of a self- 
 centred building. And corporate worship was always an 
 integral part of corporate life, so that the college chapel 
 was necessarily a very important part of its buildings. 
 So, too, the dining-hall with its kitchen and buttery were 
 as essential in a college as in a monastery. But with these 
 correspondences resemblance ceases. In the monastery, 
 the refectory and kitchen were removed as far as possible 
 from the church ; in the college, hall and chapel were 
 covered by the same roof ; college students never slept 
 in corridor-dormitories, nor forgathered in common 
 rooms the common room is a post-Renaissance institu- 
 tion in both universities nor worked in carrels in the 
 cloisters. 
 
 The earliest colleges, University, Balliol, Merton, had 
 been formed by the purchase or lease of existing private 
 houses, sometimes not even contiguous one to another, 
 but usually, as at Merton, forming an irregular group. 
 
THE COLLEGE PLAN 103 
 
 Possessing no chapel, the members used an aisle of the 
 parish church ; thus the early students of Balliol wor- 
 shipped in the north aisle of S. Mary Magdalene ; those 
 of Queens in the church of S. Peter ; the members of 
 Exeter College used S. Mildred's Church, and those of 
 Merton the church of S. John the Baptist. One of the 
 first acts of the founder of Merton College was to rebuild 
 the church of the parish ; he left the work uncompleted, 
 only finishing the choir for the use of the college (1297) ; 
 the arches of the tower were built soon after, and the 
 transepts were begun ; but they were not finished until 
 1424, while the proposed nave and aisles were never 
 carried out at all. Nevertheless, both college and parish 
 used the church in common until the middle of the 
 nineteenth century, when parochial services were discon- 
 tinued. 
 
 The chapel of Merton, with the library, built in 1377, 
 and the muniment room, which may have been one of the 
 original tenements bought by the founder, formed an 
 irregular quadrangle ; possibly the library was so planned 
 as to complete the square ; at any rate this, the Mob 
 Quadrangle, is the oldest college quadrangle in existence. 
 
 When Wykeham came to plan the first complete 
 college in 1379, he thus had more than one precedent to 
 guide him ; there was the quadrangular plan, which was 
 clearly as well suited to a college as to a monastery, and 
 into which the buildings of Merton had naturally grouped 
 themselves ; and there was the detached tenement 
 arrangement, which had been found convenient in the 
 older colleges, since a senior member could be made 
 responsible for the discipline in each separate house. 
 
 There was already in existence a type of building which 
 combined both these arrangements ; it was the mediaeval 
 
104 THE GRAMMAR OF ARCHITECTURE 
 
 inn, which was formed of separate rooms or groups of 
 rooms built round a central courtyard, from which all 
 were approached ; the rooms on the ground floor opened 
 on to the yard, while those of the upper story gave on 
 to a balcony which ran at that level round the four sides 
 of the square. It was in such courtyards that the early 
 plays were performed. 
 
 The plan of the ' Golden Cross ' in Cornmarket still 
 suggests the original arrangement, and there is an inn 
 at Dorchester the ' George ' in which a part of the 
 balcony yet remains in the yard, and of which the gate- 
 way, too, is probably original. 
 
 This quadrangular grouping of separate tenements was 
 also the plan of the mediaeval hospital (i.e. almshouses) 
 as we see at Ewelme. 
 
 It is evident that the college plan, as we have it at New 
 College, is not directly derived from any one source, but is 
 simply a convenient arrangement, of mixed parentage. 
 
 Chapel and hall being lofty buildings, it was convenient 
 to make them in a single block and to cover them with 
 a single high roof, running the whole length of one side 
 of the quadrangle ; the other sides were occupied by the 
 rooms of the members. Each 'staircase' represents one 
 of the halls or houses of which the earlier colleges were 
 composed. But it is only in comparatively recent times 
 that any member, except the heads of the college, enjoyed 
 the luxury of a room or rooms to himself. Originally 
 each set of rooms consisted of one large apartment with 
 a fireplace and two big windows, and opening out of 
 it two, three, or four small studies, each lighted by a small 
 window. The men occupying the set used the large 
 room as a common sitting-room by day, and as a sleeping- 
 room by night. The senior man slept on a tall fixed 
 
THE COLLEGE PLAN 105 
 
 bed, beneath which the truckle beds of his camerales 
 (chums) were pushed out of the way in the daytime. 
 In the grouping of the windows of the older colleges this 
 original arrangement can usually be traced, even though 
 the partitions between the studies have been knocked 
 out to form a single large bedroom. 
 
 The origin of the plan adopted by Wykeham for the 
 chapel of his college is a disputed point. It was formerly 
 supposed that it was suggested by the unfinished chapel 
 of Merton, which was found to be very convenient, the 
 choir for worship, the transept for lectures in divinity, &c. 
 But it has lately been held that the indebtedness is on 
 the other side, that Wykeham's idea of a college chapel 
 was seen to be so suitable for its purpose that the dons of 
 Merton decided not to carry on the work farther than the 
 transepts. However that may be, the plan of Wykeham's 
 chapel, like that of his college, remained the type for all 
 later builders. 
 
 THE ELEVATION 
 
 We have seen that even the plan of a mediaeval church 
 was largely conditioned by the roof-building skill of the 
 founders. The majority of the Saxon churches seem to 
 have consisted of a nave and chancel only, covered either 
 by a single timber roof resting on the side walls, or 
 separately by two such roofs at different levels. This 
 simple arrangement was adopted in small churches at all 
 periods, e. g. at Elsfield, Besilsleigh, and the Hinkseys. 
 The span of a timber roof is, of course, limited by the 
 length of the beams available. Where a wide area had to 
 be roofed, it was necessary to divide it longitudinally 
 into three parts, and to roof each part separately. Inner 
 
io6 THE GRAMMAR OF ARCHITECTURE 
 
 walls, carried by two rows of arches, supported a medial 
 roof, and the aisles were covered by lean-to roofs sloping 
 up from the top of the outer walls to the bottom of the 
 inner. The nave, therefore, depended for its light mainly 
 upon the windows in the aisle walls, which still, in many 
 churches, are the principal source of illumination. 
 
 It is customary to speak of the ' dim religious light J in 
 our old churches as if their builders had aimed at pro- 
 ducing a sense of mystery in the brooding shadows. 
 The truth is otherwise ; the dim light inevitably resulted 
 from the system of roofing. Moreover, it was the constant 
 endeavour of mediaeval builders to increase the amount 
 of light in their churches. 
 
 In order to provide the nave with an independent 
 source of light, the walls which carried its roof, and were 
 themselves carried by the arcades, were raised sufficiently 
 high above the aisle roofs as to allow of the insertion of 
 a range of small windows known as a clear-story. In 
 some churches, e.g. Holywell, this is an improvement 
 of later builders, but in large churches it is usually a part 
 of the original plan. 
 
 Where the church was on a grand scale, the height of 
 the aisle roof was such that it sloped up to a line consider- 
 ably above that of the crowns of the nave-arches. The 
 clear-story wall carried by them had, therefore, to be 
 raised so that the sills of its windows were above that line. 
 This left a band of blank wall between them and the 
 arches of the arcade. To save material, to lessen the 
 weight upon the arches, and to get rid of this blank wall, 
 it was pierced, or rather replaced, by a row of small 
 arches carried by the nave arcades and carrying in turn 
 the wall of the clear-story above. Thus, in a great 
 mediaeval church, three stories can be distinguished, the 
 
FIG. 29. TYPICAL GOTHIC ELEVATION 
 
io8 THE GRAMMAR OF ARCHITECTURE 
 
 nave arcade, the clear-story, and, between them, the 
 triforium, often called the blind-story because, since its 
 arches are below the level of the aisle roof, it transmits no 
 light to the nave. Outside, only two stories are visible, in 
 the windows of the aisle and those of the clear-story above. 
 
 In our own cathedral, which is the only local church 
 built with a triforium, a strange modification of the usual 
 plan has been adopted. The nave-arches rise to the level 
 of the sills of the clear-story windows, and carry the 
 clear-story wall without the intervention of triforium- 
 arches. These, which are hence rendered structurally 
 useless, are inserted in the tympana of the great arches 
 of the nave, and are carried by lower arches springing 
 from corbels half-way up the piers. This remarkable 
 arrangement appears to have been originally tried at 
 Romsey, but was there abandoned in favour of the usual 
 plan. The Oxford builders also departed from it in 
 building the upper stories of the tower. 
 
 In many churches the nave receives a large part of its 
 light from the west end of the building. In Norman and 
 Early Gothic times, when windows were small, the west 
 wall was pierced with several openings, as is the case at 
 Iffley. In later work one great window, often the largest 
 in the building, took the place of the group of smaller 
 ones ; the west window of S. Peter's, inserted in the 
 older Norman wall in the fifteenth century, is a good 
 example. Where there was a western tower its arch was 
 raised to a great height so as to cut off none of the light 
 entering from the west. 
 
 The other end of the nave received a certain amount of 
 light from the windows of the chancel, and especially from 
 the eastern wall, which, like the western, was pierced either 
 with several small windows, as at S. Michael's (c. 1220), 
 
THE ELEVATION 109 
 
 or with one large one, as at S. Peter's (1410). Where 
 the chancel arch was low, as in early work, and cut off the 
 light from the east, or where, as was sometimes the case, 
 its tympanum was boarded up, a window was cut through 
 the wall above it, and above the level of the chancel roof. 
 Such a window may be seen at Great Milton. 
 
 CHAPTER II 
 
 THE VAULT 
 
 IT was the great ambition of mediaeval architects to 
 roof their buildings with stone, partly because it was more 
 architectural and more consistent with the rest of the 
 edifice, but mainly because a vaulted building was thus 
 rendered fire-proof ; for fires were frequent in an age 
 when houses were built of wood and thatch, and were 
 destructive when they could only be extinguished by the 
 summary method of pulling down the burning building 
 with a hook kept for the purpose in the parish church. 
 
 Almost all the few examples of Saxon vaulting are to be 
 found in crypts, where the surrounding earth prevented 
 the arched roof from spreading, and so obviated the diffi- 
 culties of providing abutment. 
 
 The Normans, of course, were more enterprising, and 
 all their important buildings were vaulted more or less 
 completely. The aisles were so planned that their width 
 was equal to the distance between the nave piers. It was 
 thus possible to divide them into squares, covering each 
 square (or bay) with a groined vault springing from the 
 four angles. Two of the four thrusts of the vault were 
 taken by the piers of the nave, their oblique pressure 
 being neutralized by the vertical weight of the clear-story 
 
i io THE GRAMMAR OF ARCHITECTURE 
 
 wall ; the other two were received at corresponding points 
 on the wall of the aisle, which was made very massive 
 to withstand them. Norman crypts were similarly cross- 
 vaulted by division into squares, and slypes or passages 
 were ceiled by means of barrel vaults. 
 
 But the problem of spanning the great medial space 
 of the nave with a stone roof baffled generation after 
 generation. Two difficulties, apparently insuperable, were 
 involved in it : the nave was wider than the aisles, and, 
 therefore, wider than the spaces between its piers ; so each 
 of its bays was an oblong and not to be roofed by cross- 
 vaulting. And then, even if a high vault could be con- 
 structed, how could its thrusts be prevented from forcing 
 apart the clear-story walls, and so bringing all to the 
 ground ? Attempts were made to ceil the nave with 
 a barrel vault, 1 but, in order to resist its thrust, the 
 clear-story walls had to be made so massive that the arches 
 below were overweighted. Other solutions were tried in 
 England, France, and Germany, but a really satisfactory 
 one was not reached until the end of the twelfth century. 
 That solution brought the Gothic style into existence ; or, 
 more correctly, it marks the passing of Norman architec- 
 ture into the first of the stages of that known as Gothic. 
 
 It was the progress made in vaulting the aisles that 
 brought into sight a practicable method of vaulting the 
 nave. In constructing a groined vault, a wooden frame- 
 work was necessary, upon which the sections of the tunnels 
 could be formed ; when the sections were completed the 
 centring was removed, and they * locked ' themselves by 
 mutual pressure. But very early in the twelfth century 
 somebody, somewhere, had discovered that, if merely a 
 skeleton framework of the intersecting curves of the groins 
 1 Cf . the middle archway of the Clarendon Press. 
 
FIG. 31. VAULT OF CHANCEL, S. PETER'S 
 Showing Norman ribs with enrichments. 
 
THE VAULT 113 
 
 were made, stone arches could be formed upon it, and the 
 four curved surfaces between them could be filled in one 
 at a time with a single section of centring used in succession 
 for each. The arches forming the skeleton of the vault 
 have been well named * ribs ' ; and the sectional method 
 was a great advance on the Roman system of building the 
 vault as a whole, since it resulted in a great saving of 
 centring, the planks for which were hard to come by 
 before the age of saw-mills ; indeed, the difficulties of 
 constructing even the centring for a vault of great span 
 would have been almost insuperable by any but the skeletal 
 method. 
 
 The reader will appreciate the advantages of the new 
 plan if he compares the groined vault of the crypt of 
 S. Peter's with the ribbed vault of the chancel. The 
 crypt had to be divided into fifteen small squares, and 
 a corresponding number of piers were necessary at the 
 angles. Then a complete set of cross-vault centring was 
 constructed, and used for each square in succession. 
 But since pillars would have been very inconvenient in 
 the chancel above, it was divided into two great squares, 
 so large that it would have been very difficult to construct 
 complete sections of centring for cross-vaults, and instead 
 of attempting to do so, the builders simply made two 
 intersecting arches of wood, built the diagonal ribs of 
 the eastern vault upon them, moved them to the western 
 bay and repeated the process, and then, with one section of 
 centring, filled in each web successively. The development 
 of the new idea, of ribbed vaulting, is nowhere more 
 completely illustrated than in the roofing of our own 
 Cathedral. It is probably not too much to say that the 
 student, having the successive stages of progress side by 
 side for comparison, may learn more of the history of 
 
 785 n 
 
ii4 THE GRAMMAR OF ARCHITECTURE 
 
 vaulting in half an hour well spent in the Cathedral 
 precincts than in a week of visits to isolated examples. 
 
 The study of the development of vaulting is the study 
 of progressive changes in the curvature, the section, and 
 the number of the ribs employed. The building of the 
 
 FIG. 32. VAULT-RIB : PRIMITIVE TYPE 
 
 FIG. 33. VAULT-RIB : TYPICAL NORMAN 
 
 Cathedral was begun at the east end (c. 1160), and the 
 earliest vault is that of the south choir-aisle. Here the 
 ribs are typical Norman ; besides the diagonal ribs 
 marking the groins of the cross-vault, there are transverse 
 arches over the aisle defining the limits of each square or 
 ' bay ' of vaulting, east and west ; all four arches are very 
 heavy and massive ; but the transverse ribs are of the 
 primitive type, square in section ; the groin-ribs are 
 
THE VAULT 115 
 
 semicircular, and therefore lighter in comparison. In 
 order to make all the ribs rise to the same level in the 
 crown of the vault the builders were obliged to give the 
 transverse arches a stilted or horseshoe form ; for since 
 the height of a round arch is always half its span, they 
 
 FIG. 34. VAULT-RIB: TRANSITIONAL 
 
 FIG. 35. VAULT-RIB : EARLY ENGLISH 
 
 would not otherwise have reached the level of the wider 
 diagonal arches. In vaulting the north choir-aisle the 
 same expedient was adopted. The lighter ribs of the 
 transept aisle show that the builders had begun to realize 
 that they had been wasting material in constructing 
 massive arches to support the thin shell of vault between 
 them, but they still know of no better way of bringing 
 arches of unequal span to the same level than that of 
 stilting the narrower. 
 
 H 
 
ii6 THE GRAMMAR OF ARCHITECTURE 
 
 Then came the true solution, the discovery that arches 
 of varying span can all be made of the same height if 
 they are formed by intersecting arcs of two circles instead 
 of being struck from a single centre ; in other words, 
 if they are made with pointed instead of semicircular 
 heads. The round arches of the transept aisle are the 
 last of that form in the Cathedral, but which was the 
 first of the new is a more open question. 
 
 The vault-ribs of the north aisle of the nave resemble 
 those of the transept aisle in section, but the transverse 
 arches have pointed heads instead of the clumsy stilted 
 form. The most easterly of these may be the first 
 pointed arch in the building. It is, however, more 
 probable that the new form was first adopted in building 
 the arches of the tower. The transept is so much narrower 
 than the nave that its two tower-arches, north and south, 
 could only reach the level of the eastern and western 
 ones by being made with pointed heads. The builders 
 would hardly have planned an oblong tower unless they 
 had designed these arches from the first. 
 
 The south aisle of the nave was the last to be vaulted ; 
 the ribs are lighter, are moulded into a pear-shaped form, 
 and their surface is relieved with narrow bands called 
 fillets. We shall treat of mouldings in another chapter, 
 but it may be said here that the moulded ribs of this 
 aisle, pear-shaped and filleted, are extremely good 
 examples of the early days of Gothic. 
 
 In the work of the next generation, in the vault of the 
 Lady Chapel (c. 1210), and of the Chapter House (c. 1220), 
 the result of the introduction of the pointed arch is at 
 once evident : while the ribs have decreased in size, 
 they have increased in number. This development was 
 inevitable now that any number of arches could be brought 
 
FIG. 36. VAULT-RIBS IN CATHEDRAL 
 
 The rearmost show the mouldings of the twelfth century, the foremost 
 those of the fourteenth, and the ribs springing from the central column 
 those of the thirteenth. 
 
ii8 THE GRAMMAR OF ARCHITECTURE 
 
 to the same level irrespective of their span ; so here, 
 besides the diagonal and transverse ribs we have wall- 
 ribs to north and south, so that the vault rests upon 
 six arches, and its thrusts are brought down to its four 
 angles by means of twelve ribs, springing in threes from 
 the piers or vaulting-shafts. The bays of the Chapter 
 House afford a particularly good illustration of the advan- 
 tages arising from the elastic proportions of the pointed 
 arch ; they are narrow oblongs, and so the wall-arches are 
 acutely pointed, while the transverse arches are obtuse, 
 and the diagonals are semicircular ; yet all rise to a 
 common level in the crown of the vault. 
 
 Now that the ribs were becoming so thin it was not 
 easy to mitre them at the vertices ; the workmen seem 
 to have discovered this in building the vault of the 
 Lady Chapel, for in vaulting the Chapter House they 
 adopted the simpler plan of fitting them into a common 
 keystone or ' boss ' at their junction in the crown of the 
 vault. 
 
 The Gothic treatment of bosses serves admirably to 
 illustrate the true architectural principle of beautified 
 construction as opposed to the false one of applied 
 ornament ; the boss is simply a workman's device to 
 simplify his task, but the artist-workman of early Gothic 
 days made it at the same time a means of beautifying the 
 vault it knit together. How art and science can be united 
 in architecture may be seen in the bosses of the Chapter 
 House, or in the wonderful pendants of the choir, which 
 are simply elongated bosses. 
 
 The next step, which is illustrated in the vault of the 
 Latin Chapel, was to connect the bosses at the crowns 
 of the rib-arches by means of short horizontal ribs called 
 ridge-ribs, so binding them all together by a sort of 
 
izo THE GRAMMAR OF ARCHITECTURE 
 
 continuous keystone. Thus a further advance was sug- 
 gested ; it was to break up the thrusts of the vault still 
 further by introducing intermediate ribs (tiercerons) spring- 
 ing from the angles and rising to the ridge-ribs midway 
 between the vertices of the main arches, as in the vault 
 of the cloisters and in Exeter College Chapel and the 
 Gateway of Merton. The final step, seen in the vault 
 of the choir, was to bind all the ribs together by short 
 horizontal braces called liernes, which were often so 
 disposed, e.g. in the roof of the Proscholium, as to 
 produce star-shaped patterns in the crown of the vault ; 
 hence the term stellar vaulting for this kind of work. 
 The simplest type of lierne vault is that under the 
 Warden's Lodge at Merton. 
 
 By the introduction of tiercerons and lierne ribs, the 
 skeleton became a network, with interstices so small as 
 to be bridged over with single flat stones ; thus centring, 
 except for the main ribs, became unnecessary. Also, 
 the vault was reduced in thickness to a mere shell, 
 exercising very little thrust ; - and so piers, arches, and 
 vaulting-shafts could be made correspondingly lighter, 
 and with less cost for material. It therefore became 
 possible even for the village builders to gratify themselves 
 and their little society by building a stone roof. But 
 most village churches were already sufficient in size for 
 the needs of the community. That is why we find so 
 many vaulted porches added to older churches in the 
 fourteenth and fifteenth centuries. 
 
 All through the fifteenth century and well into the 
 sixteenth (in Oxford, indeed, in the seventeenth) lierne 
 vaulting remained popular. But after the middle of the 
 fourteenth century it shared its popularity with a rival 
 form, to which itself had given birth. This was the 
 
122 THE GRAMMAR OF ARCHITECTURE 
 
 beautiful fan-tracery vaulting, which was invented at 
 Tewkesbury and perfected in the cloisters of Gloucester 
 (c. 1400). Most of our local examples are of late date; 
 the earliest is the vaulted passage to All Souls Chapel 
 (c. 1440), and the finest is the vault of the stairway to 
 Christ Church Hall, built in 1640. 
 
 Having thus briefly traced the development of vaulting 
 from the groined roof of S. Peter's crypt to the fan- 
 traceried vault of the Christ Church stairway, we must 
 go back to our original twelfth-century problem of how 
 to poise a stone roof on the clear-story walls of an aisled 
 nave. As we have seen, one part of the problem had 
 been solved before the end of the century, when the intro- 
 duction of the pointed arch had made it possible to vault 
 an oblong bay as easily as a square one. Before the end 
 of the century, too, the builders had realized that the 
 stability of a vault depended not so much upon the 
 general mass of the wall as upon the provision of sufficient 
 abutment at the points where its thrusts were concentrated 
 at the springing of the rib-arches. These points, there- 
 fore, they fortified with heavy piers, and by building 
 masses of solid masonry against the walls of aisles at the 
 points where the vault-thrusts were received. But clear- 
 story walls could not be so fortified ; if they were made 
 massive enough to withstand the thrusts of the vault, 
 they would overload the arches below ; neither could 
 buttresses be built against them, because the ground was 
 already occupied by the aisles of the nave. 
 
 The builders of the Cathedral seem to have thought 
 that the pointed arch had solved all their vaulting 
 problems ; and when they had finished the aisles they 
 began, as their vaulting-shafts show, upon the ceiling of 
 the transept. Then they must have realized that though 
 
THE VAULT 123 
 
 they could put a vault up, they could not find means to 
 keep it up, and so they abandoned the idea and made 
 their high roofs of timber. Meanwhile, in France, had 
 they but known it, the problem had been solved. Opposite 
 the points in the clear-story walls where the thrusts 
 of the vault were concentrated, the French architects 
 built solid masses of masonry outside the walls of the 
 aisle ; and, by means of stone bars above the aisle roof, 
 transmitted to these the thrusts of the vault-arches. 
 Thus, if the external buttresses were heavy enough to 
 stand upright against the outward thrusts transmitted 
 by the flying buttresses, the vault would remain stable 
 till the stones crumbled from sheer decay. 
 
 We have in Oxford no ancient example of a high vault 
 sustained by flying buttresses ; by the fifteenth century, 
 when the choir of the Cathedral was vaulted, arches had 
 become so depressed, and their reduced thrusts were so 
 broken up among the network of ribs, that the mass 
 of the clear-story wall alone provided sufficient abutment. 
 The choir-vault, too, springs from so low a level that its 
 outward thrusts are met and partly neutralized by the 
 inward push of the vaulting-arches of the aisles. Mr. 
 Bodley's church of the Society of S. John the Evangelist 
 affords a good modern illustration of the use of the flying 
 buttress ; the thrusts of the stone arches of the nave 
 roof are transmitted over the aisles to external buttresses, 
 which, loaded as they are with pinnacles, are massive 
 enough to annihilate them. The west wall of the 
 Wesleyan Chapel in Walton Street is also fortified with 
 small flying buttresses to enable it to support the thrusts 
 of the arches of the nave arcades. The architect wished 
 to build a vestibule against this wall, and, therefore, was 
 obliged to move his buttresses back and transmit the 
 
1 24 THE GRAMMAR OF ARCHITECTURE 
 
 pressure to them by means of flying buttresses above its 
 roof. 
 
 The reader may perhaps wonder why I have devoted 
 so much time to the subject of vaulting, when stone- 
 roofed buildings are so rare ; still more, why I have 
 treated of flying buttresses, having none but modern 
 illustrations. In the succeeding chapters I hope to make 
 it clear that almost every detail of an unvaulted building 
 finds its origin in a vaulted one, and owes its form to its 
 relation with a vaulted roof. And it is impossible to 
 appreciate Gothic architecture unless we can realize that 
 it is a system of building in which the thrusts of arches 
 are concentrated and absorbed at fixed points by inert 
 masses ; the buttress, not the pointed arch, is the essential 
 feature of Gothic : and the function of the buttress is 
 best realized when it is seen as an external mass of masonry 
 whose sole raison d'etre is to receive transmitted thrusts. 
 
 That in all our modern Gothic we have so few vaulted 
 roofs shows how little the revivalists understood the 
 principles of the style. The timber roofs they copied 
 are really confessions of weakness in the original builders 
 they lacked either the means or the skill to bring 
 their work to its logical conclusion. Buttresses that 
 abut no arches are as futile as niches that contain no 
 images : but the modern builder continues to produce 
 both. 
 
 CHAPTER III 
 
 THE BUTTRESS 
 
 THE history of the buttress might almost be said to 
 be the history of Gothic architecture. It, and not the 
 pointed arch, is the touch-stone of Gothic. 
 
THE BUTTRESS 
 
 125 
 
 The buttress had been used by the Romans, but was 
 not trusted by them nor by the Romanesque builders 
 who followed them ; they preferred to rely on the massive 
 solidity of their walls. But in a pure Gothic building, 
 e. g. the Divinity School or the Sainte Chapelle as 
 represented by the Chapel of Exeter College, the walls 
 
 : 
 
 M 
 
 tn 
 
 FIG. 39. NORMAN BUTTRESS 
 
 are screens, not supports ; they might be made entirely 
 of glass, or even removed altogether, and the roof would 
 still remain poised aloft, balanced upon its shafts and 
 buttresses. 
 
 The essential function of the buttress is to resist oblique 
 thrusts ; but it serves another purpose, that of binding 
 and unifying the wall in which it is built ; this is, of 
 course, its only purpose in an unvaulted building. Its 
 use greatly economized material, both by relieving the 
 
iz6 THE GRAMMAR OF ARCHITECTURE 
 
 walls of weight and by allowing them to be made thinner 
 with no sacrifice of strength ; for it is obvious that a 
 Gothic wall two feet thick, with buttresses projecting 
 three feet, is as strong as an unbuttressed Romanesque 
 wall of double its thickness. 
 
 The Saxons used no buttresses ; they strengthened 
 their walls by ' long and short work ' long stones, 
 alternately vertical and horizontal, built into the wall as 
 in the angles of S. Michael's tower. The Norman builders, 
 as may be seen at Iffley, tied their rubble walls at intervals 
 with broad flat bands of squared stones built vertically 
 into the wall and projecting slightly from its face. But 
 these are clearly not designed to take the thrusts of the 
 roof ; they are therefore not true buttresses, and their 
 origin is rather to be found in the pilasters with which 
 the Greek and Roman builders broke up the blank 
 faces of their walls ; they may be compared with the 
 pilasters on the front of the Ashmolean Museum. The 
 somewhat similar strips set in pairs at the eastern 
 angles of the church are typical of the work of the 
 next generation (c. 1200) ; the thin shaft on the angle 
 is an easily recognized mark of the transitional buttress. 
 They may possibly have been built when the chancel 
 was lengthened, and for the purpose of supporting the 
 new vault ; if so, they must be regarded as true but- 
 tresses, the earliest examples we have ; but that their 
 united weight was insufficient for its purpose is clear 
 from the deep buttresses added in the fourteenth 
 century, and the modern pair set against the east wall 
 (Fig. 40). 
 
 But the thirteenth-century builders soon discovered 
 that what was wanted in a buttress was mass, and 
 they increased their projection accordingly, e.g. in the 
 
FIG. 40. IFFLEY CHURCH, EAST END 
 
128 THE GRAMMAR OF ARCHITECTURE 
 
 buttresses of the Chapter House (c. 1220). They saw, too, 
 that walls might now be made thinner, but they did not 
 
 FIG. 41. BUTTRESS, S. MARY MAGDALENE (c. 1337) 
 
 realize how much thinner, and so the walls of the Chapter 
 House are still very massive. Neither did they appreciate 
 the fact that depth, not width, was the important dimen- 
 
THE BUTTRESS 129 
 
 sion in a buttress, and at first material was wasted in 
 both these ways. 
 
 The early practice of setting small buttresses in pairs 
 at the angles of buildings, at the east ends of Iffley and 
 Cowley Churches, for example, soon gave way to the 
 more economical method of building one large buttress 
 diagonally against the corner, as in the south aisle of 
 S. Mary Magdalene. Soon, too, the thirteenth-century 
 builders realized that the thrusts of the roof passed 
 gradually outwards and downwards, and so they increased 
 the projection of their buttresses from the top downwards, 
 i.e. they built them in diminishing stages from the 
 ground to the roof. In the first half of the century the 
 tops were made to slope into the wall at or below the 
 line of the eaves ; but later on, as in Merton Choir 
 (c. 1297), they were carried above the wall and crowned 
 with a gablet. This not only threw off the rain, but 
 by its vertical weight assisted the buttress to resist oblique 
 thrusts. The builders of the south aisle of S. Mary 
 Magdalene Church (c. 1337), appreciating this, loaded 
 their buttresses with pinnacles. The two examples last 
 mentioned illustrate once more the artist spirit in Gothic 
 work ; the buttress, in essence a mere inert lump of 
 heaviness, is so treated that one knows not whether to 
 admire more its structural efficiency or its artistic beauty ; 
 and that though the peculiar beauty of the fourteenth- 
 century buttress, the niched image, has been ravished 
 by brutal ignorance. Beautiful as they remain, the central 
 object on which the supreme care of the carver was 
 lavished has disappeared, and only its setting remains for 
 our admiration. There was almost as much labour in 
 a single perfect fourteenth-century buttress as there is 
 in the Martyrs' Memorial. 
 
 785 T 
 
130 THE GRAMMAR OF ARCHITECTURE 
 
 The buttresses of the next period, e.g. in New College 
 Chapel (1386) or the Divinity School (1450), show even 
 more science, but less art. It had become clear that thrusts 
 
 FIG. 42. BUTTRESS, DIVINITY SCHOOL (c. 1450) 
 
 were to be met by projection, i.e. depth, in the buttress, 
 and so we have the deep thin type seen in the buildings 
 quoted. But here the architect is saving labour as well 
 
THE BUTTRESS 131 
 
 as material ; and what is worse, he is stinting design. 
 Any one could cover a buttress with panels like those 
 of the Divinity School. These buttresses were built to 
 support the most ingenious vault yet seen in England, 
 or in the world ; yet they proclaim that the end of Gothic 
 art was in sight ; they are covered with cheap ornament, 
 and though science may clean your carpets for a pittance, 
 cheap art never has been nor ever will be. 
 
 When the Classic styles were revived, the Roman 
 principle of disguising abutment caused the buttress to 
 be abandoned ; the last buttresses to be built in Oxford, 
 until the revival of Gothic, were the huge ugly masses 
 in Exeter College Gardens, piled up against the Divinity 
 School by Wren to support its walls against the weight 
 of the books in the library above. 
 
 CHAPTER IV 
 
 THE ARCH 
 
 THE function of the arch is to carry weight, usually 
 that of the mass of wall above an opening ; this weight, 
 of course, exerts a vertical pressure, but the arch transmits 
 it to its supports obliquely, tending, like a bent spring, 
 to force them apart ; they will only remain immovable 
 when the vertical pull of gravity on their mass is stronger 
 than the oblique thrust of the arch ; in other words, the 
 thrust of the arch is met by the inertia of a dead weight, 
 which is termed the abutment of the arch. The weight 
 of the abutment is really set in balance against the weight 
 carried by the arch. 
 
 But if two equal arches meet at a common springing, 
 i 2 . 
 
132 THE GRAMMAR OF ARCHITECTURE 
 
 as in a nave arcade or on the piers of a bridge, the tendency 
 of the one to push over its support is nullified by the 
 opposing thrust of the other, and the resultant of the 
 two forces is a vertical pressure needing no abutment ; 
 the mass of the support at this point is therefore of no 
 importance, since it has only to resist compression. In 
 the nave arcade of S. Giles's Church, for example, the 
 only points where abutment is needed are at the western 
 springing of the western arch, and the eastern springing 
 of the eastern one. The thrust at the first point is taken 
 by the dead weight of the tower wall, and that at the 
 other by the wall of the chancel. This is the structural 
 purpose of the western tower ; where it does not exist, 
 as in S. Mary's Church, its place is taken by two massive 
 buttresses set against the western wall in line with the 
 arcades. The massive piers of central towers are similarly 
 explained ; they have not only to support the weight 
 of the tower, but to resist the thrusts of the arches that 
 abut upon them, in which, of course, they are helped by 
 the dead load above them. 
 
 The arches of a bridge neutralize each other's thrusts 
 where they meet upon the piers in mid-stream ; it is only 
 upon the banks that they require abutment, and this is 
 afforded by the mass of roadway forming the landward 
 approach to the bridge. 
 
 In Roman and in Romanesque buildings down to the 
 twelfth century the semicircular arch was employed ; its 
 weakest point is its crown, i. e. the very point where the 
 greatest weight falls upon it. The pointed arch, on the 
 other hand, is strongest at its vertex, and for this reason 
 was occasionally used by the Normans very early in the 
 twelfth century, as being the superior weight-carrier. 
 
 Much has been written about the * discovery ' of the 
 
THE ARCH 133 
 
 pointed arch, which is often said to have been invented 
 by the Gothic builders, or borrowed by them from the 
 Saracens. But it had been known and appreciated as 
 a weight-carrier long before, and indeed, like the semi- 
 circular form, is probably prehistoric. So long as the 
 round arch served all their purposes the Romanesque 
 builders preferred that form, perhaps because it was 
 easier to construct (since the method was traditional), 
 perhaps because they considered it more beautiful : 
 certainly it was the form of the great arch type, the heaven 
 that encompassed all. But pointed arches are by no means 
 uncommon in Norman work ; they may be seen, for 
 example, in the ornamental arcade on the south wall of 
 S. Peter's Church (Fig. 66), where their formation by 
 intersecting semicircles is clearly shown. 
 
 The builders of the late twelfth century no more 
 discovered the pointed arch than the Romans did the 
 round one ; what they did discover was its application 
 to the problems of vault construction with which they 
 were faced, and which its elastic proportions enabled 
 them to solve. Once admitted into the vault, it spread 
 to the rest of the building. But the round arch persisted 
 long after Norman times ; it may be seen, for instance, 
 in the porch of Cuddesdon Church in a gateway of 
 about 1300. The doorways of that church, too, are round 
 arched, though all their details are Early Gothic, and the 
 tower-arches of the same date are pointed. 
 
 The typical arch of the thirteenth century (c. 1190- 
 1270) is of the lancet form, acutely pointed, and formed 
 by the intersection of two equal circles, each having 
 its centre outside the circumference of the other. Then, 
 as windows increased in size, the equilateral arch became 
 general ; it was formed by the intersection of two equel 
 
i 3 4 THE GRAMMAR OF ARCHITECTURE 
 
 circles, the centre of each being in the circumference of 
 the other (Euc. I, 3). This is the typical arch of 
 'Decorated' work (c. 1270-1350), though it was fre- 
 quently used at every period. 
 
 FIG. 43. LANCET ARCH FIG. 44. EQUILATERAL ARCH 
 
 Early in the fourteenth century the ogee arch was 
 discovered, and soon came into general use. It is formed 
 of arcs of four circles, and was suggested by the line of 
 double curvature formed by the combinations of circles 
 in the geometrical window-tracery of the late thirteenth 
 century. It really * discovered itself ' in the window, and 
 
 BL 
 
 FIG. 45. DROP ARCH FIG. 46. OGEE ARCH 
 
 the window was always its realm, for it is too weak for use 
 as a weight-carrier. It is essentially a decorative arch, and 
 is common over tombs and niches, e.g. over a tomb-recess 
 at Waterperry (Fig. 122), and in the niches of S. Mary 
 Magdalene buttresses. A small doorway in Witney Church 
 has an ogee arch, and in some districts, e.g. Northampton- 
 shire, window-arches of this form are not uncommon. 
 In the heads of the lights the ogee arch is almost universal 
 
THE ARCH 
 
 135 
 
 in the fourteenth and fifteenth centuries, and its curves 
 dominate the tracery from about 1310 almost to the end 
 
 FIG. 47. FOUR-CENTRED ARCH FIG. 48. 
 
 LATE FOUR-CENTRED 
 ARCH 
 
 FIG. 49. FOURTEENTH-CENTURY DOORWAY, WITNEY 
 
 of the century. It was one of the few Gothic forms that 
 had any attraction for Wren, who used it in his doorway 
 to the Divinity School (Fig. 17), and in the curves of his 
 
136 THE GRAMMAR OF ARCHITECTURE 
 
 dome at Christ Church ; he seems also to have altered 
 the arch of the great window over the gateway to make 
 it correspond with the dome above. 
 
 Another ornamental form, in the thirteenth century, 
 was the trefoil arch, formed from three circles having 
 their centres at the angles of an equilateral triangle. This 
 had been occasionally employed by the Romanesque 
 builders. There is a trefoil-headed doorway at Charl- 
 bury, but the form is seldom used for large arches ; it 
 is oftenest seen in the heads of recesses, e.g. in the 
 sedilia at Iffley, or in the heads of window-lights, as in 
 the east window at Northmoor (Fig. 90). 
 
 In the latter half of the fourteenth century the builders, 
 who were experimenting with fan-tracery vaulting, dis- 
 covered a form of arch that simplified the construction 
 of their vaults. They found that this new and convenient 
 arch which they had hit upon empirically was formed 
 of arcs of four circles, i. e. like the ogee arch, it was struck 
 from four centres ; but, unlike the ogee form, it was 
 sufficiently strong to be used as a weight-carrier. As it 
 was employed everywhere in subsequent work it came to 
 be known as the four-centred arch. It was natter and 
 wider than the earlier forms, and so it was welcomed by the 
 builders, who were being called upon for larger windows. 
 It is so characteristic of the work done in the building 
 revival that followed the Wars of the Roses that it has 
 been called the Tudor arch ; it is the arch of S. Mary's 
 Church and Christ Church; but it is also the arch of 
 New College, built a century before Bosworth (Fig. 47). 
 
 A debased form of the four-centred arch is very 
 characteristic of Late Gothic work : the arch becomes 
 a gable curving only at its springing, as in the head of 
 the doorway of Kettel Hall (Fig. 48). From this to the 
 
FIG. 50. INTERIOR OF S. MARY'S (c. 1488) 
 
138 THE GRAMMAR OF ARCHITECTURE 
 
 flat-headed windows of Stuart Gothic, as seen in the Old 
 Schools, is but a short step. This ' gabled ' arch is antici- 
 pated in a curious fourteenth-century window in the 
 chancel at North Hinksey, and in a similar one in the 
 transept chapel at Cumnor. It should be said here that 
 though the building system of the Middle Ages was 
 essentially an arcuated one, yet the lintel never entirely 
 disappeared ; the simplest way to roof a small opening 
 is to bridge it with a long stone, and doors and windows 
 were so treated at all periods. In the south wall of Cowley 
 Church (Fig. 97), for instance, every window has a flat 
 head ; yet those in the chancel are of the thirteenth 
 century, one in the nave is of the fourteenth, and another 
 is of the late fifteenth. In humble domestic architecture, 
 too, where timber-work was the rule, doors and windows 
 must certainly have been lintelled. 
 
 The Priest's Door of S. Thomas's Church illustrates 
 a form that was in use in the thirteenth century ; two 
 stone brackets or corbels project from the angles of the 
 opening, thus lessening its span, and a long stone rests 
 upon them to form the head of the door. This construc- 
 tion has been called the corbelled arch or the shouldered 
 arch, but it is really a corbelled lintel, and in no sense 
 an arch, since it exercises no thrust. It was used in 
 building Carnarvon Castle and the other fortresses of 
 Edward I hence it is sometimes known as the Carnarvon 
 arch and it is conspicuous in the modern front of 
 Balliol College. 
 
 In the fourteenth century segmental arches are some- 
 times found in the window-heads. In eighteenth-century 
 work this form is common, and is employed, for example, 
 in the windows of the Clarendon Building (Fig. 21). 
 
 Having discussed thus briefly the function and forms 
 
FIG. 51. PRIEST'S DOOR, S. THOMAS'S (c. 1220) 
 
140 THE GRAMMAR OF ARCHITECTURE 
 
 of the arch, we pass on to consider the method of its 
 construction. And here the application of what has been 
 called the biological method to the study of architecture, 
 the attempt to approach the subject rather from the 
 scientific than the artistic side, will lead to some interesting 
 conclusions. We shall discover, for instance, that the 
 beauty of the famous west doorway of Iffley results 
 inevitably from the Romanesque method of arch-building, 
 and is not, like a picture or a statue, a beautiful creation 
 of a single mind ; we shall find that the graceful clustered 
 shafts of the Lady Chapel were not, as used to be believed, 
 suggested by twin-trunks or grouped pine-trees ; that 
 they were never, in fact, conceived as ideas by any one, 
 but shaped themselves inevitably to a form best suited 
 to the construction of the arches they carry. They are 
 beautiful : but their beauty, like that of the horse or 
 the honeysuckle, arises from the perfect fitness of their 
 structure to its purpose. 
 
 When an arch is to be constructed, the builders have 
 first to make a wooden framework of the required size 
 and shape, and to place it upon the supports designed 
 for the arch. Upon this temporary structure they build 
 the first ring of the arch, usually of wedge-shaped stones 
 called voussoirs. These are allowed to set, and then they 
 themselves form a supporting arch for other rings to be 
 built above them. When the Roman builder made an 
 arch, he had such massive stones at his disposal that he 
 could build the first ring the full thickness of the complete 
 arch ; e. g. if his arch spanned an opening in a wall four 
 feet thick he could obtain voussoirs of that width ; and, 
 moreover, he had appliances by which such great stones 
 could be lifted into position. But the Romanesque work- 
 man had no such resources ; he had to construct an arch 
 
THE ARCH 
 
 141 
 
 in a four-foot wall with small stones, raised by rude 
 tackle ; and he had not the necessary timber for the 
 massive centring necessary for Roman arches. Therefore 
 he was driven to devise a new plan : setting his frail 
 centring in the middle of the archway, he built upon it 
 
 FIG. 52. ARCH CONSTRUCTION, ROMAN METHOD 
 
 with small voussoirs one thin arch ; upon this as a centring 
 he laid a second ring, composed not of single voussoirs, but 
 of stones set in pairs, so that their edges projected over 
 the face of the first ring on both sides, producing a broader 
 surface on which a third and still wider ring could be 
 constructed. By increasing the number of rings he could 
 make his arch of any desired thickness. But, the soffit or 
 underside of an arch so constructed would not be flat 
 as in a Roman arch cut straight through the wall, but 
 
142 THE GRAMMAR OF ARCHITECTURE 
 
 broken up and ' recessed ' as many times as there were 
 orders, i. e. rings. In order to give unity to this com- 
 position, the Romanesque builders chamfered off the angles 
 of each ring and carved ornaments on the sloping surface 
 so produced. The inner orders of the Chapter House 
 doorway are so treated ; but the two outer orders have 
 
 FIG. 53. ARCH CONSTRUCTION, MEDIAEVAL METHOD 
 
 had their angles rounded into the form of a bold roll 
 called a torus. At Iffley and S. Peter's the inner orders 
 are similarly treated, and the torus on the two outer ones 
 is ornamented with beak-heads. In Late Norman work 
 the angles of the orders are generally moulded, i. e. rounded 
 off, instead of being chamfered and ornamented. In the 
 chancel arch of the Cathedral, for instance, the outer 
 orders are moulded, the edges of the inner ones being 
 left with their original square edges. 
 
THE ARCH 143 
 
 In Gothic work the method of construction was the 
 same, but the arch was relieved by mouldings and not 
 by carved ornament as in Norman times. In work of the 
 early thirteenth century one can usually trace the original 
 rings of the arch, since the mouldings are cut on the 
 chamfered faces of the angles ; but in later work the unity 
 of the arch is complete, and the suites of mouldings are 
 independent of the orders in which it was constructed. 
 But if these are sometimes not easily to be detected 
 in the arch, the evidence of their existence is generally 
 to be found in the pier, to which we have now to turn 
 our attention. 
 
 CHAPTER V 
 
 THE PIER 
 
 THE pier is the columnar mass of masonry that supports 
 the springing of an arch and receives its thrust. It thus 
 differs from the Classic column both in function and in 
 structure ; the former supported the vertical pressure 
 of a lintel and consisted of a monolith of marble or of 
 drums of stone ; the latter, having to resist an oblique 
 pressure, must be more massive, and is therefore built 
 of courses of masonry. The Romans, who introduced 
 the arch, refused to recognize its complement, the pier, 
 and sinned against the first law of architecture, that 
 design should express construction, by disguising it with 
 columns carved on its face. The Romanesque builders 
 had not the skill, even if they had the wish, so to disguise 
 construction, and in their work the pier is unmasked, as 
 the buttress is in that of their successors. 
 
144 THE GRAMMAR OF ARCHITECTURE 
 
 The Roman arch consisted of a single ring ; its support- 
 ing pier was therefore a simple square or oblong prism 
 (Fig. 52). But the Romanesque and Gothic arches were 
 built in rings, each ring or order requiring its own proper 
 
 FIG. 54. ARCH AND PIERS, S. MARY'S, A.D. 1488 
 
 support (Fig. 53) ; moreover, in vaulted buildings the ribs 
 of the vault also came down to the pier, and each needed 
 its proper shaft from which to spring. Therefore the post- 
 Roman pier, like the post-Roman arch, is composite ; is a 
 combination of grouped members, each supporting an order 
 
THE PIER 145 
 
 of the arch or a rib of the vault. The only alternative is a 
 great circular pier supporting all the orders and ribs, and in 
 early Romanesque work this form is not uncommon, since 
 it was easy to construct. But the truth-loving Teutonic 
 mind demanded a clear correspondence between the 
 load and the carrier, and in later work each order in the 
 arch has its proper support in the pier. Oxford, however, 
 even in the twelfth century, appears to have been strongly 
 conservative, clinging then, as it did until the other day, 
 to older methods ; so in the Cathedral we find all the 
 orders of a recessed arch brought down to one large 
 circular pier. But in the next century logic prevailed 
 even here, and the composite structure of the arches of 
 the Lady Chapel is acknowledged in the clustered columns 
 of the supporting piers. 
 
 The later Norman builders sometimes emphasized this 
 correspondence by employing detached shafts to carry the 
 orders of an arch. This practice suggested a means of 
 increasing the contrasts of light and shade by the use of 
 shafts of a different stone. The magnificent tower-arch 
 of Iffley owes much of its effect to the massive shafts 
 of black marble that carry the orders. The use of 
 marble shafts became common in the next century in 
 the greater churches ; the poorer builders could not 
 afford them, for the supply in England was practically 
 limited to the marble of Purbeck, and the cost of transport 
 was too heavy. The labour and expense of turning many 
 separate columns made detached shafts the exception in 
 parish churches. The only instance of the use of Purbeck 
 marble in Oxford is in the small shafts of Prior Sutton's 
 tomb in the Cathedral (c. 1300). Detached shafts of stone 
 are found in a few buildings, but only in small arches, as 
 in the south doorway of S. Giles's. By the end of the 
 
 785 K 
 
FIG. 55. THIRTEENTH-CENTURY PIERS, S. PETER'S 
 
FIG. 56. THIRTEENTH-CENTURY PIER, HOLYWELL CHURCH 
 
148 THE GRAMMAR OF ARCHITECTURE 
 
 thirteenth century builders everywhere had become con- 
 tent to do what had been done in the piers of the Lady 
 Chapel and of S. Peter's, viz. to carve engaged shafts on 
 a central core. 
 
 In parish churches at all periods the early economical 
 plan, of making a single simple pier carry all the orders 
 of the larger arches, was the general rule. Such piers were 
 either circular or hexagonal 1 until the fifteenth century, 
 when they were usually octagonal. In Holywell Church 
 the thirteenth-century piers of the south arcade, like 
 those of S. Giles's, are cylindrical ; those of the fifteenth 
 century on the north are octagonal in plan ; so too are 
 the late fourteenth-century piers of S. Mary Magdalene 
 and the fifteenth-century ones of S. Michael's. But 
 by the fifteenth century, even in village churches, the 
 builders had become such skilful craftsmen that they were 
 usually able to make the form of their pier correspond 
 to the orders of the arch. So, alike in the arches of 
 Wykeham in New College Chapel, and in those of the 
 village craftsmen at Eynsham, Beckley, and Ewelme, the 
 orders of the arch rest upon corresponding members in 
 the supporting pier. We may mourn the failing art in 
 fifteenth-century work, but we have to set against it the 
 increase in constructive science. This is seen as well in 
 the mass of the piers as in their form. The late fourteenth- 
 century piers at Islip or Clifton Hampden, for example, 
 do the same work as the twelfth-century piers in the 
 opposite aisles, with half the expenditure of material. 
 
 The pier, like the column, has three parts, the capital, 
 shaft, and base. The capital is a bracket which, accom- 
 modating its shape and size below to the pier and above 
 to its load, enables the former to carry the arch of a wall 
 1 Rarely octagonal. 
 
FIG. 57. TOWER PIERS, MELTON COLLEGE (c. 1300) 
 
150 THE GRAMMAR OF ARCHITECTURE 
 
 wider than itself ; in its most elementary shape the 
 capital would therefore be an inverted truncated cone. 
 But no mediaeval builder would have been content to 
 leave it thus ; the artist spirit transformed it into a grace- 
 ful bell, sometimes decorated with carved foliage, some- 
 times ringed about with mouldings. 
 
 The Romanesque builders, confronted with the task 
 of making a pier support a load wider than itself, placed 
 upon it a cube of stone equal in width to the thickness 
 
 FIG. 58. CUSHION CAPITAL, FIG. 59. SCALLOP CAPITAL 
 CASSINGTON (c. 1140) (c. 1160) 
 
 of the arch, and rounded off the lower angles, so pro- 
 ducing the form known as the cushion capital. This may 
 be seen in its simplest form in the crypt of S. Peter's. 
 The first step in its elaboration was the refining of its 
 broad round faces by grooves like those in a scallop-shell, 
 which increased the play of light and shade ; so was 
 produced the scallop capital seen in the south doorway 
 of Cowley Church. The cushion capital was also orna- 
 mented with rude figure-sculpture carved in relief. Those 
 of the south doorway at Iffley are world famous, and 
 illustrate the dominating ideas of the warlike age in which 
 
THE PIER 
 
 15* 
 
 they were cut. As Norman passes into Gothic these 
 semi-barbaric ornaments are abandoned, and the capital 
 is decorated with leaf-forms, not apparently derived from 
 any specific plant, but suggested, perhaps, by the acanthus 
 leaves of the Corinthian column. The capitals in the nave 
 and choir of the Cathedral are very representative of the 
 
 FIG. 60. CAPITAL IN CATHEDRAL (c. 1180) 
 
 work of this period, and the student who will go a little 
 farther afield will find other good examples in the capitals 
 of the nave arcades of Appleton, Haseley, and Islip, in 
 doorways at Holton and Cuddesdon, and in the chancel 
 arch of Elsfield. The capitals in Chesterton Church and 
 some of those at Appleton, show plainly the stages of 
 evolution from the scalloped form of the twelfth century 
 
152 THE GRAMMAR OF ARCHITECTURE 
 
 to the foliaged capital of the thirteenth. One peculiarity 
 of this transition period is the presence of a volute of 
 tightly-curled foliage at the angles of the capital ; it is 
 very well illustrated at Elsfield. 
 
 Logic demanded that the load and its support should 
 be clearly differentiated, and so the line of contact between 
 arch and capital is marked by a member called the 
 abacus. In Romanesque work, as in Roman, the abacus 
 is usually a square tile-like stone. On the Continent this 
 form was retained in Gothic work, but in England, after 
 the Gothic style was firmly established, the abacus was 
 made circular. Moreover, though in Norman work its 
 plan was sometimes round, its upper edge was always 
 square. But after the twelfth century, not only was the 
 abacus circular in plan, but its edges were rounded off 
 so that it showed no angles either in plan or elevation 
 (Fig. 62). An almost infallible test for the work of the 
 last quarter of the twelfth century is the presence of 
 a square abacus on a capital decorated with foliage. 1 
 
 The rounding of its uppermost member gave to the 
 Gothic capital a bell-like contour, which it never wholly 
 lost, even when after the thirteenth century it was made 
 polygonal in plan. But it was most beautiful when most 
 bell-shaped, i.e. in the thirteenth century. Then, in 
 a typical capital such as those in the Chapter House 
 leaves spring from the necking where the capital rests 
 on the pier, thrust themselves upward close pressed 
 against the bell, and then, as if their upward growth were 
 stayed by the abacus, curl over and break into fronds of 
 foliage. These leaves have all the beauty of natural 
 
 1 A thirteenth-century capital on the south-east respond of 
 Bicester Church is the only English exception known to the 
 writer. 
 
FIG. 61. PIER OF CHANCEL ^RCH, ELSFIELD (c. 1200) 
 
 Showing typical transitional capital with square abacus 
 
 and volute of foliage, and water-holding base. 
 
154 TH E GRAMMAR OF ARCHITECTURE 
 
 leafage, but it is an abstract beauty, not a copy of a parti- 
 cular plant. No natural leaf could so combine dissimilar 
 qualities give strong support to the abacus, and then, 
 relaxing its stiffness, wreathe it round with wind-blown 
 lightness. As craftsmanship increased, the workman tried 
 his hand on natural foliage, producing, for instance, the 
 lovely bosses of leafage of oak, maple, vine, ivy, and 
 bryony seen on S. Frideswide's shrine ; but the frail 
 leaves, however beautiful, could not be made to express 
 support ; instead of lending strength to the capital, they 
 clung to it as parasites : they were beautiful ornaments, 
 not beautified construction. 
 
 FIG. 62. CAPITAL (c. 1220) 
 
 Still, by the end of the thirteenth century the natural- 
 istic capital had everywhere superseded the stiff-stalked 
 type. Though inferior in design to the latter, it required 
 great skill in its execution, and is therefore seldom seen 
 in parish churches. S. Giles's, S. Peter's (Fig. 55), Holy- 
 well (Fig. 56), have stiff-stalked foliage on the caps of their 
 thirteenth-century piers and shafts, but no parish church 
 in or near Oxford could command the services of workmen 
 able to carve naturalistic capitals ; after the thirteenth 
 century, common men had to be content with mouldings 
 instead of leaf-carvings. 1 
 
 1 The moulded capital was always common even in the thirteenth 
 century, as in the south arcade of Holy well, the north arcade of 
 S. Giles's, and the chapel of S. Peter's. 
 
THE PIER 155 
 
 Builders who could afford it, used the naturalistic 
 capital all through the fourteenth century. But the 
 carving steadily deteriorated ; the marvellous leafage on 
 S. Frideswide's shrine was copied from leaves growing 
 in Oxfordshire lanes in the summer of 1289 or there- 
 abouts ; a hundred years later all such fidelity to nature 
 had disappeared ; conventional leaves were again in 
 fashion ; but this time not because strength as well as 
 beauty was wanted in the foliage, but because conven- 
 tional leaves were easier to cut. The thirteenth-century 
 foliage had both beauty and structural expression ; that 
 of the fourteenth had the one without the other ; but 
 
 FIG. 63. CAPITAL, MERTON COLLEGE CHAPEL (c. 1280) 
 
 the foliage on fifteenth-century capitals has neither ; it 
 sticks on the bell like a dead parasite, as may be seen on 
 examining the capitals in the Cathedral cloisters. 
 
 Though the characteristic Gothic capital is decorated 
 with foliage, the moulded capital was common at all 
 periods. The development of Gothic mouldings will be 
 considered in a later chapter, when it will be seen that 
 a capital can be dated by the mouldings as certainly as by 
 the leafage. It is here sufficient to say that the piers of the 
 south arcade of Holywell, of the two arcades of S. Giles's, 
 and of the chapel-arches of S. Peter's, have moulded 
 capitals of the thirteenth century ; the tower-arches of 
 Merton (Figs. 57 and 63) and the vaulting-shafts of the Latin 
 Chapel (Fig. 36) have moulded capitals of the fourteenth; 
 
156 THE GRAMMAR OF ARCHITECTURE 
 
 and the piers of S. Mary's nave, and of the ante-chapels 
 of New College and All Souls, show on their capitals the 
 characteristic mouldings of fifteenth-century work. It 
 
 FIG. 64. PIER, S. MARY'S, A.D. 1488 
 
 will be seen in the examples last mentioned that the 
 abacus of the fifteenth century is a concave-sided polygon 
 in plan, a circumstance which is alone sufficient to dis- 
 tinguish a Late Gothic capital from those of earlier date. 
 In order to lower the centre of gravity in his scaffolding, 
 
THE PIER 157 
 
 the modern builder plants his poles in tubs of clay ; so 
 a mediaeval builder sought to give stability to his piers 
 by setting them upon substantial bases of stone. 
 
 The Norman piers, like the columns of the Doric 
 Order, were themselves so massive as to need little 
 weighting at their feet. The pier was placed upon a 
 square plinth, and a bold roll-moulding concealed the 
 circular line of junction ; in the angles left between the 
 circle and the square a spur of ornament was often 
 placed, especially in late work. In the transition period 
 two plinths are superimposed, and the roll-moulding 
 upon them is hollowed out to receive the pier, so that a 
 deep groove runs round the foot of the latter. The base 
 characterized by this channel is known as- the ' water- 
 holding ' (Fig. 6l) ; it is well seen in the Cathedral and 
 in the buttress-shafts at the east end of Ifney Church. 
 The water-holding hollow was retained in Gothic work 
 well into the thirteenth century, for instance, in the bases 
 of the piers in the south arcade of Holywell Church 
 (Fig. 56) ; but the square plinth did not long survive 
 the square abacus. Henceforward, the plan of the base 
 corresponds more or less to that of the shaft. 
 
 As the pier grows thinner in the fourteenth and 
 fifteenth centuries bases increase in size, and especially 
 in height. The fifteenth-century bases to the piers of 
 S. Mary's, for instance, are four feet high. Wooden 
 seats, which before had been luxuries confined to the choir 
 and chancel, were now introduced into the nave, hiding 
 from view the low bases of earlier days. So the bases now 
 built were carried up above the seats, at the level of which 
 most of the mouldings were placed, the lower courses 
 being left plain, as in the piers of the north arcades of 
 S. Mary Magdalene and Holywell. The bases of the 
 
158 THE GRAMMAR OF ARCHITECTURE 
 
 former illustrate a common characteristic of the fifteenth- 
 century type ; their circumference is greatest, not at the 
 level of the floor as in earlier bases, but at the level of 
 their lowest mouldings, which are worked on a projecting 
 course of masonry overhanging a plain pedestal. The 
 bases of the shafts (Fig. 69) in the jambs of the doors and 
 
 FIG. 65. FIFTEENTH-CENTURY BASE, S. MARY'S, 1488 
 
 windows in Merton transepts (c. 1425), and especially 
 those in the north doorway (Fig. 80), are fine examples 
 of the fifteenth-century pedestal base. 
 
 The piers of the Lady Chapel in S. Mary Magdalene 
 Church have neither capitals nor bases, and other examples 
 of this illogical arrangement may occasionally be met 
 with for instance, in the north arcade at Clifton Hamp- 
 
THE PIER 159 
 
 den. The explanation is generally that the arches were 
 not built upon piers first set up to receive them, but 
 were cut through an already existing wall, the piers 
 representing portions of it left to support them. The 
 S. Mary Magdalene Chapel was built by the Carmelite 
 Friars of Beaumont, in the second quarter of the fourteenth 
 century, against what was then the south wall of the 
 church. At first it was a separate building, approached 
 by the lane still known as Friars' Entry. Afterwards it 
 was desired to throw it into the church, and to save labour 
 the old wall was left standing, merely being pierced with 
 arched openings. The thirteenth-century builders at 
 S.Giles's had had to do the same thing when they built the 
 north aisle of that church, but they did it in a more archi- 
 tectural way. They cut arches through the upper part of 
 the wall, as may be seen from the traces of the original 
 clear-story windows, and supporting them on temporary 
 props, built piers with capitals and bases, no doubt using 
 a strip of the old wall as a central core for each pier. 
 
 CHAPTER VI 
 
 THE ROOF 
 
 WE have said that if the reader could imagine a typical 
 Gothic building with its walls and their windows and 
 doorways all removed, the skeleton then remaining would 
 represent the essential parts of Gothic construction. With 
 this fundamental structure, the vault with its supports of 
 arch, pier, and buttress, we have hitherto been dealing in 
 this section on the grammar of architecture. One other 
 adjunct to the skeleton remains to be considered before 
 we pass on to discuss the walls with their openings, 
 
160 THE GRAMMAR OF ARCHITECTURE 
 
 The Romans had roofed their more important buildings 
 with groined vaults of such massive strength that their 
 arches could carry upon their crowns a flat covering of 
 cement or flagstones, forming at once a roof and a terrace. 
 The ambition of mediaeval builders was to build, more 
 Romanorum, with stone throughout ; but though at last 
 they rivalled the Roman in skill, they never possessed his 
 resources of material, and they were always obliged to 
 protect the external face of their vaults by a mask of tiles 
 on a framework of rafters. 
 
 A few roofs do exist composed entirely of stone, that 
 of the thirteenth-century Muniment Room at Merton 
 College is an example ; but that is not a vaulted roof, 
 but simply one in which a framework of stone rafters 
 supports the flags. Even the Renaissance architects were 
 usually content with wooden roofs though we have 
 small examples of stone in the domes of Queen's College 
 Gateway and All Souls Cloisters; and the modern archi- 
 tect, when called upon to build a stone roof, generally 
 supports it upon iron girders, as in the roof-playground 
 of the South Oxford school. 
 
 Since a timber covering must be constructed, whether 
 a vault is beneath it or not, it is obvious that the builders 
 of parish churches had generally to be content with that 
 part of the complete roof which kept out the weather, 
 especially as this part was much the easier and cheaper 
 to construct. But it must be repeated that Gothic 
 architecture was developed, not in the parish churches, 
 but in the masons' schools of abbeys and cathedrals ; 
 and that forms originally invented in relation to vaulting 
 problems were co-pied in parish churches, even though 
 these were unvaulted. 
 
 The function of the roof being to throw off rain and 
 
THE ROOF 161 
 
 snow, its structure must obviously bear reference to 
 prevalent conditions of climate. In the sunny climate 
 of Greece, when a separate roof-mask had to be built 
 to protect a perishable ceiling, it was made with a very 
 low pitch, in harmony with the prevailing horizontal line ; 
 its triangular end thus formed a low pediment and not 
 a gable. This was often filled with figure-sculpture in 
 relief, as in the eighteenth-century pediments of Queen's 
 College. Early roofs in our own rainy island were of 
 ' very steep pitch ; the gable angle being sometimes one 
 of 45 in a thirteenth-century building. Norman roofs 
 were also steeply sloped ; though very few now remain, 
 their original pitch may usually be seen on the face 
 of the central tower. The original twelfth-century 
 roof of the Cathedral was particularly acute, as may be 
 seen from the well-marked angle on the tower-faces 
 (Fig. 27). 
 
 The eastern gable of S. Peter's Church, too, shows the 
 pitch of the original Norman roof (Fig. 66) ; its small 
 window is very interesting ; it must once have lighted 
 a room above the vaulting of the chancel, used perhaps 
 by the priests or the sacristan. The pitch of the existing 
 roof is of the same date as the porch, which also has 
 a room above its vaulting, and thither the fifteenth- 
 century occupant must have transferred his quarters. 
 The little window in the eastern gable of IfHey (Fig. 40) 
 still lights a space between the vault and the timber- 
 roof, which is accessible from the tower. 
 
 The roof of Merton Choir, built in the last years of 
 the thirteenth century, is an excellent example of the 
 high-pitched Early Gothic roof; the gables of S.Giles's 
 Church also retain the sharp angles of the thirteenth 
 century. 
 
 785 L 
 
THE ROOF 163 
 
 Steeply-pitched roofs, however, are usually modern 
 restorations of the original form ; most of our existing 
 ancient roofs are of the fifteenth century, built when the 
 low four-centred arch dominated the building. The 
 low-pitched roofs then put up are not to be regarded 
 as a mere fashion, significant of lower ideals in their 
 builders. They are certainly not so beautiful as the high 
 gables and shaly slopes of earlier roofs with their long 
 ridges cutting the skyline like a distant mountain-range, 
 but they are to be explained and justified on constructional 
 grounds ; the case against the fifteenth-century artist is 
 clear enough without any evidence from his roofs, which, 
 like all his construction, tells rather on the other side. 
 
 In the first place, owing to an improved system of 
 roof-drainage, presently to be described, the high pitch 
 was no longer a necessity, and it was in accordance with 
 all Gothic tradition to use no more material than was 
 necessary ; labour was lavished freely to beautify what 
 was useful, but material was never wasted for the mere 
 sake of effect on the eye. Long beams and rafters were 
 scarce and dear, short ones common and cheaper ; more- 
 over, their use reduced the superficial area of the roof ; 
 hence the fifteenth-century preference for roofs of low 
 pitch. Where an old roof had to be rebuilt, the argument 
 from economy is still stronger. When, for instance, the 
 fifteenth-century builders found it necessary to pull 
 down the high-pitched Norman roof of S. Peter's Church, 
 they must have had plenty of sound timber to use again. 
 Beams and rafters decay most where they rest upon the 
 walls and at their junction with each other, i.e. at their 
 ends ; if the rotten ends are sawn off, shorter lengths of 
 sound wood remain to be used again in a roof of lower pitch. 
 
 But there is yet another reason for the flatness of Late 
 
 L2. 
 
1 64 THE GRAMMAR OF ARCHITECTURE 
 
 Gothic roofs. The development of lead-mines in the 
 fifteenth century enabled the builder to cover his rafters 
 with sheets of that metal instead of tiles. But lead on 
 a steeply sloped roof would tear itself away from its 
 fastenings by its own weight and by its expansion and 
 contraction in sun and frost. Finally, while ruddy tiles 
 and lichened shingles are picturesque in themselves, and 
 add to the charm of a building, a lead roof can never be 
 beautiful, and on artistic grounds alone the builders were 
 justified in sinking its pitch and hiding it from view by 
 a decorated parapet running round its eaves. 
 
 Mention of the parapet leads us to a consideration of 
 the way in which the drainage of the roof was effected. 
 The system in general use in Norman and Early Gothic 
 times was invented when the first roof was constructed 
 in the world, and is in use now in every thatched barn 
 and cottage. It is the simple plan of dripping eaves. 
 In order that the drippings shall not* run down the walls 
 and so in time destroy them, the eaves of the roof are 
 made to project above the top of the wall, so that the 
 water shoots off and sinks into the ground some distance 
 from the foot of the building, leaving walls and founda- 
 tions free from damp. 
 
 The early roof was therefore wider than the space it 
 covered, and so it rested, not immediately upon the wall 
 itself, but upon a projecting course of masonry built 
 out upon blocks of stone, forming a row of brackets known 
 as a corbel-table. The corbels, of course, were generally 
 made an ornamental feature, as in the fine corbel- 
 table of S. Peter's (Fig. 66) ; those of the Cathedral eaves 
 (Fig. 27) are plainer, and those of Iffley show that the 
 Norman budilers never completed their work, since a 
 carved face here and there indicates their intention of 
 
THE ROOF 165 
 
 beautifying their corbels as they had done their capitals. 1 
 An Early English corbel-table, carrying pointed arches 
 which support the projecting eaves, may be seen in the 
 tower of the Cathedral below the springing of the spire 
 (Fig. 27). 
 
 It will be seen, however, that most Norman corbel- 
 tables now carry, not the roof, but a projecting parapet, 
 usually of the fifteenth century. The roof, moreover, 
 instead of being wider than the building, is narrower, 
 and its rafters rest on the inner half of the thickness of 
 the wall ; there are therefore no dripping eaves. 
 
 The drainage of the roof of Merton Choir illustrates 
 the new system which was invented in the latter part 
 of the thirteenth century. It will be seen that a thin 
 parapet was built, partly upon a corbel-table of grotesque 
 heads, partly upon the outer portion of the thickness of 
 the wall ; inside the parapet runs a deep gutter, lined 
 with lead to prevent the water from soaking into the wall ; 
 and sloping up from the inner side of this is the roof, 
 its rafters resting on the remaining surface of the top 
 of the wall. At intervals along the length of the gutter 
 the parapet is pierced with projecting spouts of stone 
 called gargoyles even these are ornamental as well as 
 useful and the whole drainage of the roof is concentrated 
 at these points, and by the force of its own volume is shot 
 out clear of the building. 
 
 The gutter was a great improvement upon the dripping 
 eaves, and after the thirteenth century it became the rule 
 in the roofing of all large buildings and in the rebuild- 
 ing of older roofs. That is why the twelfth-century 
 
 1 Similarly, a single richly ornamented window on the south face 
 of the tower shows that the belfry-story, fine as it is, falls short of 
 the beauty of the original design. 
 
166 THE GRAMMAR OF ARCHITECTURE 
 
 corbel.-table of S. Peter's now carries a fifteenth-century 
 parapet. But the corbel-table was usually dispensed with 
 in new work, and parapet, gutter, and roof were carried 
 by the top of the wall. 
 
 It is possible that the new plan of parapet and gutter 
 was suggested by the curtain-walls of military fortifica- 
 tions. The top of the city walls, which were built or 
 rebuilt in the reign of Henry III, forms a terrace along 
 which a watchman could walk, or on which archers could 
 take post if the city were attacked. A thin curtain-wall, 
 loop-holed and embattled, built upon its outer edge, 
 protected the defenders from the arrows of the enemy. 
 This may well have suggested to the builders of Merton 
 the idea of a parapet safeguarding a narrow space on the 
 top of the wall, which should be at once a gutter and 
 a path by which men could walk round the building for 
 purposes of defence or for inspection and repair of the 
 fabric. There can be no doubt that the church-builders 
 of the Middle Ages built always with mingled ideas of 
 worship and war. Their iron-bound doors with massive 
 locks and beam or bar from jamb to jamb are alone a proof 
 of this ; the narrow windows of early times, high-set in 
 the walls, and the iron stanchions of later days, have the 
 same significance. The tower might be a landmark to 
 the traveller in the pathless waste, a home for the bells, 
 and an abutment for the nave-arches, but it was also 
 a fortress for the villagers, and with its narrow winding 
 stair and unscalable faces it was an impregnable refuge 
 in time of need. The tower of S. Michael's was at once 
 a part of the church and of the city fortifications ; that 
 of New College was both a muniment-room and a bastion. 
 Like Durham Cathedral, every church tower was 
 Half church of God : half castle. 
 
168 THE GRAMMAR OF ARCHITECTURE 
 
 The military value of a parapet must have had a good deal 
 to do with its popularity. Along the eaves of the low aisle 
 roofs it was usually little more than an ornamental rail, as 
 in the fourteenth-century south aisle of S. Mary Magda- 
 lene, but on more defensible buildings, and especially on 
 towers, it was always embattled for archery. Of course, 
 the builders appreciated too, the beauty of a battlemented 
 outline ; one sign of failing artistic sense in late Gothic 
 is the introduction of diminutive battlements in all sorts 
 of incongruous situations, as for instance, along the tran- 
 soms of the large fifteenth-century window of S. Peter's. 
 The presence of decorative battlements among the orna- 
 ments of the Martyrs' Memorial is a similar artistic 
 blunder, and is the less excusable because the artists of 
 the original crosses were never guilty of such confusion 
 of ideas. 
 
 Throughout the Middle Ages the church, the hall, and 
 the tithe barn, were roofed in the same way by the same 
 workmen ; there is a thirteenth-century barn at Coxwell 
 which is roofed by means of piers and arches, and but- 
 tressed exactly like a church, except that the work is 
 plainer. The cottage was a one-storied building with 
 a thatched roof, upon the open beams of which the fowls 
 perched at night above the heads of the sleepers ; even 
 in the town houses, as we learn from the ' Miller's Tale ', 
 the beams were open to the room beneath, and there 
 was no upper story. But with the rise of a middle class, 
 with the increase of population, and its concentration in 
 towns, storied buildings became necessary, and rooms had 
 to be planned in the space immediately below the rafters. 
 
 This affected the form of the roof, since these rooms 
 must be lighted by windows ; so the dormer- or roof- 
 window came into use ; it was at this time that upper 
 
i;o THE GRAMMAR OF ARCHITECTURE 
 
 stories were added to so many of the colleges, for 
 instance, New College, Brasenose, and Corpus ; the 
 ' cock-lofts ' of Trinity show how the dormer window 
 made rooms in a space which once had really been the 
 roosting-place of the ' cock beneath the thatch '. In 
 Flanders, where the woollen manufacture had brought 
 a wealthy trading-class into existence, house-building had 
 developed much earlier ; the curved gables in the eaves 
 of University and Oriel show the influence of Flemish 
 domestic architecture on English builders. 
 
 In the seventeenth century the roof was often hipped, 
 i.e. it was made without gables, sloping from its ridge 
 to four eaves instead of two ; there is a good roof of this 
 kind on the Alms-houses in S. Clement's, built in the 
 reign of Queen Anne; a common eighteenth-century form 
 was the Mansard roof, invented by a Frenchman of that 
 name ; it gave more head-room in the attics ; the roof 
 of the Radcliffe Infirmary (1770) is a good example. 
 
 Upper stories in houses made stairs necessary ; the 
 upper floors in mediaeval buildings, where they existed, 
 e.g. in keeps and in church towers, were approached by 
 means of newel-stairways, narrow and winding in a spiral, 
 or by external steps like those that give access to a modern 
 stable-loft ; such inconvenient arrangements could serve 
 no longer, and so the splendid Elizabethan and Jacobean 
 staircases came into existence. At this time, too, the 
 boards that formed the floor of the upper story were 
 hidden from below by means of lath and plaster ceilings. 
 These at first were decoratively treated, being divided 
 into panels like a fifteenth-century vault and embossed 
 with painted ornament in relief. We have nothing to 
 compare with the ceiling of the Combination Room of 
 S. John's College, Cambridge, but there is a fine Eliza- 
 
THE ROOF 171 
 
 bethan example at All Souls, and seventeenth-century 
 examples in houses in Holywell and Magdalen Street. 
 In Georgian days the plain whitewashed ceilings of to-day 
 became general, and even the carved timbers of earlier 
 open roofs were hidden by them, as may be seen at 
 Wood Eaton and in the hall of Jesus College. The open 
 timber roof of University College Hall was hidden by 
 a plaster abomination until 1904, when its original 
 appearance was restored. 
 
 THE SPIRE 
 
 Before we leave the subject of roofs something must be 
 said of the one feature of mediaeval building construction, 
 in which material was used obviously and primarily for the 
 mere sake of producing an effect upon the eye. The single 
 Gothic luxury, and that not a common one, was the spire 
 
 And even the spire had a constructional origin and even 
 a structural use ; its germ is to be found in the low pyramid 
 that roofed the Norman tower. Few of these remain, 
 but the tower-roof of Dorchester Abbey is a reconstruc- 
 tion of the original form. Raise the height of the square 
 pyramid, and you have a spire in embryo. But the faces of 
 a square pyramid present four broad surfaces to the pressure 
 of the wind ; more angles are necessary if the spire is to 
 stand safely; an octagonal pyramid on the square tower is 
 obviously suggested. That is the form of the Gothic spire. 
 
 It is almost incomprehensible that Oxford, always 
 distrustful of new notions, should have produced what is 
 often said to be the earliest spire in England. Grafted 
 on a building twenty years behind its time, S. Frideswide's 
 spire is so * stumpy ' in its proportions that it must be 
 regarded as an experiment ; its windows prove that it 
 belongs to the first half of the thirteenth century, and 
 
172 THE GRAMMAR OF ARCHITECTURE 
 
 it is often assigned to the early years of that period. Of 
 that, however, there is no evidence either way, and it is, 
 a -priori, more likely that it is a cautious Oxonian experi- 
 ment made in the middle of the century when spires had 
 been built in other places. Otherwise, we have to believe 
 that men in Oxford who built a Romanesque cathedral 
 when their contemporaries at Ripon were building in 
 pure Gothic, were succeeded by sons who were as much 
 in advance of their time as their fathers had been behind 
 it. It is not the custom of Oxford to move so fast. 
 
 But, if our claim to possess the earliest spire in England 
 cannot be proved, we can surely boast the possession of 
 the most beautiful in ' minaret crowned S. Mary's ' : 
 
 That peerless spire 
 S. Mary's own, so fair, so free, 
 The flower of perfect symmetry. 
 
 The tower and spire of S. Mary's belong to the latter part 
 of the thirteenth century. It is sometimes said that the 
 plain stately tower was not originally built for the 
 beautiful coronal it carries. Those who hold this view 
 can know little of either the art or the methods of the 
 thirteenth century. A building then was not completed 
 by contract in a fixed time by relays of workmen, but 
 was the growth of a generation. One generation planned 
 it and commenced to build ; another finished the work, 
 as men and material could be obtained. The lower 
 windows of S. Mary's tower show that the building 
 was begun about 1265 ; the windows in the next story 
 are at least ten years later, and so are the mouldings of the 
 tower-arches. By 1280 or a little later, the tower must 
 have been ready to receive its spire, which, by its profusion 
 of ball- flowers, must have been completed about 1300. 
 
THE ROOF 173 
 
 Of course, the builders of 1265 did not foresee the 
 precise form that was to crown their work ; but their 
 great buttresses show that they were building towards 
 a spire. Moreover, the plainness of their work proves, 
 not that they were planning a humble superstructure, 
 but precisely the opposite ; the spire was to be their 
 glory, and with consummate art they designed to heighten 
 the beauty of the blossom by contrast with the stem from 
 which it sprang. It remained for later and less gifted 
 architects to cover their buildings with ornament, and to 
 produce as much monotony by its profusion as would 
 have marked its total absence. Magdalen Tower, the 
 most beautiful of Late Gothic buildings, is so because 
 its builders remembered what had been forgotten for 
 a century, and lavished all their ornaments on its top- 
 most story. 
 
 We have said that the spire is not without structural 
 purpose : it serves to weight the tower, as the pinnacle 
 weights the buttress, and to help it in resisting the thrusts 
 of arches. But its great function is the artistic one of 
 grouping around it all the parts of the building into one 
 harmonious whole. 
 
 The pinnacles at the base of a thirteenth-century spire 
 likewise fulfil both constructional and artistic purposes. 
 They not only lead the eye to where the central shaft 
 springs heavenward, but by their weight they deflect its 
 thrusts in a vertical direction. Further, the spire is an 
 octagonal pyramid supported by a square tower ; hence, 
 while four faces rest upon the tower walls, the other four 
 have to be carried by arched supports, called squinches, 
 built in its angles. The pinnacles at these points hide the 
 junction and convert tower and spire into one composition. 
 After the thirteenth century (when the advantages of the 
 
i 7 4 THE GRAMMAR OF ARCHITECTURE 
 
 parapet in facilitating repairs had been appreciated) this 
 was effected by means of a parapet round the base of the 
 spire, and so the. pinnacles were omitted. The fifteenth- 
 century spire of Kidlington is an example of the parapetted 
 type that superseded the broach spires of the thirteenth 
 century. 
 
 CHAPTER VII 
 
 THE WALL 
 
 THE wall has usually a double function ; it serves as 
 a screen enclosing a roofed space and as a support for the 
 roof itself. In Romanesque buildings every wall fulfils 
 both these functions, but in the more highly organized 
 Gothic construction the walls are relieved of weight 
 and become thin screens of masonry or rather of glass and 
 masonry, for windows occupy a large part of their 
 surface : the ' walls ' of the Divinity School, for example, 
 may almost be said to be sheets of glass. A building is 
 more or less Gothic according to the degree in which 
 the weight-bearing function is transferred from the wall 
 to the buttress; the Divinity School is, therefore, the best 
 example of Gothic construction that Oxford possesses. 
 
 The Romanesque wall was necessarily thick and massive. 
 Owing to lack of skill in the workmen, both masons and 
 toolmakers, it was built of rubble, i.e. of unhewn stones. 
 In Saxon times squared stones were so difficult to obtain 
 that they were economized to the utmost ; so in the 
 corners of the buildings, where they were necessary to 
 form the right-angled quoins, the long squared stones 
 are set upright, the shorter ones are laid horizontally 
 between them, and so is produced the ' long and short ' 
 work seen in the angles of S. Michael's tower (Fig. 10). 
 
THE WALL 175 
 
 In Gothic buildings the thickness of the wail varies 
 inversely with the pro j ection of the buttress. Thirteenth- 
 century walls are still very massive, containing so much 
 material that they are usually of rubble, though they may 
 be faced with squared stones. But, as the builders learned 
 to trust the buttress, less and less material was required, 
 and so they were able to use squared stones throughout. 
 The fifteenth-century wall even in humble churches is 
 frequently of ashlar. 
 
 But squared stones are expensive, however economically 
 they may be used, and a cheap substitute was soon sought 
 for ; it had been sought for, and found, yet earlier when 
 even the mighty eastern empires and wealthy Rome herself 
 had eked out the supply of ashlar by rectangular slabs of 
 baked clay. 
 
 But with the fall of Rome bridkmaking became a lost 
 art. The barbarians of Italy found an apparently inex- 
 haustible quarry of hewn stones in her temples and palaces; 
 those of France and England fell back on rubble ; and 
 then, as the demand for finer material revived, bricks were 
 re-invented after a lapse of a thousand years. 
 
 It was, naturally, in those flat districts where clay is 
 common and stone rare, that the builders were first 
 driven to the re-discovery. Bricks were made in Flanders 
 in the thirteenth century, and were used in the east of 
 England before the end of that period. In our own 
 district, where stone is plentiful, we have few examples 
 of early brickwork ; the hospital at Ewelme, built in 
 1440, and the porch of Sutton Courtney Church a 
 hundred years later, are almost the only local illustra- 
 tions of the art before the increase of population made 
 brick the universal material for house-building. 1 say 
 universal, because even buildings apparently made of stone 
 
176 THE GRAMMAR OF ARCHITECTURE 
 
 are in modern times composed of brick with a mere 
 facing of ashlar. The builders of Keble College were 
 more honest ; but their work unfortunately does little 
 to enforce the truth of the proverb. 
 
 An examination of early brickwork shows that, though 
 the English might be indebted to the Flemish for the 
 material, they invented their own method of using it. 
 
 In Flemish work all the courses are alike, each consisting 
 of bricks showing long and short faces alternately ; in 
 the early English examples a course of bricks all showing 
 a long face alternates vertically with another in which 
 all the bricks show their narrow ends ; this arrangement 
 is therefore known as the English bond. It is stronger 
 than the Flemish bond, but it went out of fashion in 
 the seventeenth century, and was superseded by the 
 Flemish method still followed by the modern bricklayer. 1 
 The companion art of tile-making either survived the 
 fall of Rome or was much sooner rediscovered ; tiles 
 were certainly made in Normandy before the Conquest. 
 They were used both in the roofs and in the flooring 
 of mediaeval buildings in Oxford; where the Stonesfield 
 shingles were available, we have 'stone roofs grey and old ', 
 but in the walls of village churches many a bit of red 
 roof-tile, used to level up an uneven course of rubble, 
 indicates that the modern slate roof was preceded by one 
 of tiles. The roof was so exposed to the elements that all 
 our ancient examples are of thick shingles ; but mediaeval 
 floor-tiles are common in most of our local churches. 
 They are usually about five inches square and one inch 
 
 1 Since this was written, I have seen 'English' bond in the 
 brickwork of the thirteenth-century Binnehof at The Hague and in 
 other early Dutch examples. The truth probably is that the English 
 retained the older method when the Dutch had abandoned it for 
 the new. 
 
THE WALL 177 
 
 thick. While the clay was still soft a pattern was stamped 
 upon them, and the hollow thus formed was inlaid with 
 clay of another colour ; when the tile was dry it was 
 glazed and baked, and the pattern became as imperishable 
 as the tile itself. There are fine examples in the floor 
 of Merton College Library. 
 
 We have already seen how the mediaeval wall was 
 protected from the disintegrating action of the drip- 
 pings from the roof ; but further provision was made to 
 ensure that raindrops blown against the face of the build- 
 ing should be prevented from trickling down into the 
 foundations, or from depositing sediment on the glass of 
 the windows. 
 
 Round the head of each window a little ridge of stone 
 juts out like an eyebrow and, indeed, serves the same 
 purpose ; at the level of the springing of the arch it is 
 continued horizontally along the face of the wall and 
 round the buttresses, rising into an arch again at the next 
 window. This projecting band is known as a string- 
 course ; that part of it which overarches the window- 
 head is called the hood-mould ; and when, as often 
 happens in fifteenth-century buildings, it forms a square 
 frame over an arch (Fig. 38), it is distinguished as a label. 
 From its function the string-course is often described 
 'as the drip-stone, but it is frequently found on inner 
 walls, where the term cannot properly be applied. 
 
 This, however, suggests its second purpose, viz. to act 
 as a binding-course, strengthening and unifying the wall. 
 Its value in this respect is well seen at Iffley (Fig. li), 
 where the heavy strings round the building recall the 
 binding-courses in the rubble walls of Roman Silchester 
 and Verulamium. But the string-course is an element 
 of beauty, not less than of strength ; it relieves the face 
 
 785 M 
 
FIG. 69. NORTH TRANSEPT, MERTON COLLEGE CHAPEL (c. 1420) 
 
THE WALL 179 
 
 of the dead wall, and by its strong horizontal line corrects 
 the appearance of disproportionate height to which Gothic 
 buildings are liable, giving them something of the majesty 
 and breadth that marked the older system of Greece 
 
 rfl 
 
 FIG. 70. NORMAN STRING-COURSE (c. noo) 
 
 and Rome united with their own expression of soaring 
 energy. It suggests ideals and aspirations controlled by 
 sound sense in a well-balanced mind. 
 
 The Norman string-course was always heavy. In village 
 churches it is usually a square projection with its lower 
 
 FIG. 71. TRANSITIONAL STRING-COURSE (c. 1180) 
 
 angle chamfered off (Fig. 70) ; at Iffley both upper and 
 lower angles are so treated, and the string forms a semi- 
 hexagonal projection; at S. Peter's "the square string- 
 course is carved into billets (Fig. 66). In Late Norman 
 work, as in the Cathedral, the angles are rounded off, not 
 chamfered, so that the string is semicircular in section ; 
 and in the transition period it is pared away still further 
 to form the ' keel ' moulding (Fig. 71). 
 
 The next step, which marks the Early Gothic work of 
 
 Mj 
 
i8o THE GRAMMAR OF ARCHITECTURE 
 
 the thirteenth century, is to hollow out the under side 
 of the half-round so that it becomes concave "below, 
 convex above. This, is an improvement, artistic and con- 
 
 FIG. 72. EARLY ENGLISH STRING-COURSE (c. 1220) 
 
 str'uctional ; the dark hollow strengthens the line of the 
 string-course, and the overhanging ' nose ' better throws 
 off the wet. It is well seen in the hood-moulds of the 
 lancet windows of Iffley and S. Giles's Churches. 
 
 FIG. 73. DECORATED STRING-COURSE (c. 1300) 
 
 In fourteenth-century mouldings strong contrasts are 
 avoided ; black and white gives place to grey shading. 
 So the undercut hollow is abandoned for the scroll- 
 moulding, in which a large convexity above overhangs 
 a smaller one below (Fig. 73) ; the illustration will make 
 further description unnecessary. The tower-arches of 
 Merton have the scroll-moulding to their hood-moulds. 
 
 In the fifteenth century the hollow reappears again, but 
 
THE WALL 181 
 
 it is now wide and shallow ; the upper side of the string- 
 course is often flat or but slightly curved. The drip- 
 stones over the windows of Magdalen Tower are good 
 examples of their type ; similar hood-moulds in the belfry 
 windows of Wood Eaton Church prove the tower to belong 
 to the fifteenth century, though the writers of guide- 
 books, judging only by the form of the tracery, usually 
 declare it to be Decorated in style. It may be said here, 
 and it will be proved later, that the one safe guide to the 
 date of an ancient building is to be found in the character 
 
 FIG. 74. PERPENDICULAR STRING-COURSE (c. 1450) 
 
 of its mouldings. The novice is recommended to approach 
 the study of mouldings by observing the string-courses 
 of the various periods ; each string-course is a single 
 moulding typical of those which will afterwards be found 
 in groups on arches and capitals. The deep, dark cavity 
 of the thirteenth century, the scroll-moulding of the 
 fourteenth, and the shallow hollow of the fifteenth, once 
 familiarized in the drip-stone, will be recognized at once 
 in whatever combination they may subsequently be 
 discovered. 
 
 In many buildings, and especially in those belonging 
 to the fourteenth century, the string-course is only 
 represented by the hood-moulds of the windows, and 
 terminates at the springing of the arch in a small carved 
 corbel. The east window (Fig. 88) of S. Giles's Lady 
 
182 THE GRAMMAR OF ARCHITECTURE 
 
 Chapel (c. 1260) will serve as an illustration ; the drip- 
 stone terminations are examples of a corbel very charac- 
 teristic of the thirteenth century ; from a front view it 
 resembles a buckle, and is often called the buckle-corbel ; 
 but seen sideways it has the profile of a human face, and 
 therefore it is also known as the mask-corbel. The hood- 
 moulds of the nave-arches of S. Giles's stop upon bosses 
 of foliage ; this form of corbel is often met with in 
 thirteenth-century work, and grotesque carvings are also 
 frequent. 
 
 After the thirteenth century, carved heads are very 
 common as terminations to the drip-stones of arches; 
 frequently a crowned head upon one side is balanced by 
 a mitred one on the other, or by the head of a crowned 
 woman. There can be no doubt that these were some- 
 times portraits of contemporary sovereigns and bishops, 
 e. g. those in the transepts of Merton (Fig. 69) ; but since 
 the faces of kings and queens could seldom have been 
 familiar to the village carver, they must usually have been 
 purely conventional and imaginary. To those who know 
 the fashion of the head-dress at different periods they 
 often afford valuable evidence as to the date of a building; 
 a still more valuable testimony may be found by the 
 student of heraldry in the shields of arms that are often 
 used in the fifteenth century instead of heads ; a head 
 may be imaginary : not so an armorial bearing. 1 
 
 In late work, when art was weary, these carved corbels 
 are seldom found ; the hood-mould turns at the springing 
 of the arch as if to continue as a string-course and then 
 stops abruptly. In Tudor work it often terminates in 
 a diamond-shaped panel. In the last years of expiring 
 Gothic, in the sixteenth century, the drip-stone became 
 1 At least in the Middle Ages. 
 
THE WALL 183 
 
 a plain square frame over a square window-head, as may 
 be seen in the windows of S. John's (Fig. 96), and in the 
 early seventeenth-century fronts of Wadham, Oriel, and 
 University Colleges ; at last it disappeared altogether, 
 and we find the windows of the Old Schools (Fig. 79) as 
 innocent of hood-moulds as those of the Classic buildings 
 which were already superseding Gothic. 
 
 The drip-stone was not the only device by which the 
 mediaeval builder protected his walls ; experience taught 
 him that water in his foundation was as dangerous as, 
 if more subtle, than fire in his roofs, and he learned to 
 take adequate precautions against both. The massive 
 walls of the Norman masons might defy the insidious 
 attacks of damp, but the more economical builders of later 
 days raised their thin walls upon a basement of thicker 
 masonry, and built a projecting course at the level of 
 a few feet above the ground, so that their foundations 
 were completely safeguarded from any possibility of 
 percolation from above. How much protection was thus 
 given to the foundations may be seen by any one who will 
 glance at the basement-course of University College ; 
 the projecting coping has been eaten away with rain- 
 drops ; yet all that is necessary to restore the wall to 
 its original soundness is to replace this decayed course 
 with new stone ; and in the present long vacation, of 
 1911, this has been done in a part of the building. 
 
 Basement-mouldings, which are most common when 
 walls are thinnest, i. e. in the fifteenth century, should be 
 compared with the pedestal-base in the piers of the same 
 period ; the correspondence between the projecting 
 course and the overhanging moulding previously referred 
 to is very noticeable. 
 
 Artistically, the effect of a basement-course is to give an 
 
1 84 THE GRAMMAR OF ARCHITECTURE 
 
 appearance of strength and solidity to the substructure of 
 a building; in Classic architecture the effect is obtained by 
 using large stones and emphasizing their mass by means 
 of deep grooves at the joints ; the basement or podium of 
 the Camera is so treated, and so are the lower courses 
 of the High Street front of Queen's College (Fig. 81). 
 
 The inner faces of mediaeval walls were covered with 
 a thick coat of plaster, which has usually been removed 
 by the modern restorer. In the halls of castles and manor- 
 houses they were hidden by tapestry or arras hung upon 
 a wooden framework a foot or so from the face of the wall. 
 In churches the wall-surface was made to serve as a 
 gigantic canvas on which could be painted incidents of 
 Bible history. Every inch of space was covered with 
 pictorial teaching or painted decoration. All this was 
 covered with whitewash after the Reformation, but in 
 a few churches considerable traces remain. Some further 
 account of the subjects of these frescoes will be given in 
 another chapter. 
 
 In very rich work, e. g. in tomb recesses, and in the 
 backs of sedilia, the wall-surface is unplastered, being of 
 finely-jointed masonry, and is enriched with diaper-work, 
 flower-forms sunk in the face of the stone in a chequered 
 pattern. There is much beautiful diapering on the 
 Eleanor Crosses (c. 1294), the pattern of which is repro- 
 duced in the Martyrs' Memorial. The front of the 
 Wesleyan Chapel in New Inn Hall Street (Mr. C. Bell, 
 1878) is enriched with similar work. 
 
 In the fifteenth century, when walls were usually of 
 ashlar, the sunk panels that enriched their outer faces 
 were sometimes reproduced on their inner ; thus in the 
 transepts of Merton the lower part of the wall has panels 
 sunk in its face. The early panels reproduced the tracery 
 
THE WALL 
 
 185 
 
 of the windows in blank, but the patterns of late fifteenth- 
 century panelling are much less elaborate. 
 
 In Tudor times wooden wainscoting began to be used 
 
 FIG. 75. LINEN PANELLING (c. 1500) 
 
 instead of tapestry. The walls of New College Hall are 
 lined with oak panelling given by Archbishop Warham, 
 and the wainscot of the Hall at Magdalen dates from 
 the same period, and is said to have come from Reading 
 Abbey. Oak panels of this date are carved with what is 
 
1 86 THE GRAMMAR OF ARCHITECTURE 
 
 known as the ' linen pattern ', so called from its fancied 
 resemblance to the folds in linen cloth. Towards the 
 end of the century wooden panelling came into fashion 
 in all large houses, and even in the churches ; there is 
 good Elizabethan panelling at Cumnor, for instance. The 
 panels of this date were small and square, and were 
 uncarved except for a moulding at their edges. 
 
 Wooden wainscoting remained in favour all through 
 the seventeenth century, but in the century following it 
 gradually gave place to the new fashion of covering inner 
 walls with paper. 
 
 CHAPTER VIII 
 
 THE DOORWAY 
 
 THE main doorway of a church is usually in the south 
 wall, midway in the length of the nave or aisle. This 
 doorway, as a rule, is large and ornamental, and is usually 
 protected by a porch. When it is in an aisle wall it is 
 often of earlier date than the aisle itself, and bears evidence 
 of having been moved from its original position and 
 reinserted farther south when the aisle was added ; the 
 Gothic builders seem to have greatly appreciated the 
 fine Norman doorways of earlier times, and we sometimes 
 find one in a church from which every other vestige of 
 Norman work has been swept away. 1 
 
 1 In such a case an examination of the piers of the chancel 
 arch will sometimes reveal traces of Norman origin, though the 
 arch itself may be much later ; the imposts of the chancel arch 
 of S. Peter's (Fig. 55) are a case in point. At Sutton Courtney 
 the old Norman chancel arch has been moved from its original 
 position to the south side of the nave. At Holywell, Headington, 
 and Forest Hill, the chancel arches are all that remain to prove 
 the Norman origin of the buildings. At Radley the font is the 
 only vestige of Norman days. 
 
FIG. 76. SOUTH DOOR, IFFLEY CHURCH (t. 1150) 
 
1 88 THE GRAMMAR OF ARCHITECTURE 
 
 The Norman doorway now at the west end of S. Ebbe's 
 Church has been moved again and again ; its original 
 position no man knoweth ; but successive generations of 
 builders had respected it for eight hundred years until 
 it was mutilated in our own times. 
 
 Directly opposite the principal opening is another 
 doorway in the north wall ; this is usually smaller and 
 plain in character ; 1 it seems frequently to have been 
 blocked up after the Reformation ; certainly it is the 
 exception to find it in use in village churches. In the 
 Middle Ages it was known as the devil's door the north 
 side of the churchyard was the devil's province, no old 
 graves will be found in its cold shades it was opened 
 during the baptism of infants for the convenience of the 
 evil one, who was supposed to follow the unbaptized child 
 to the very font, and not to give up hope until it had 
 been actually admitted into the safety of the fold ; then, 
 baffled, he went out into the shadows by means of the 
 door left open for the purpose. This door was also used 
 when the parishioner came to church for the last time ; 
 the body entered by it and passed through the building 
 to the graveyard on the south, in token that the way 
 to heaven lies through the church. 
 
 In town churches, where mediaeval arrangements have 
 been altered to suit modern convenience, both the doors 
 may be equally used ; in some village churches, e. g. Iffley, 
 the north door has become the principal entrance, since 
 most of the village now lies on that side ; but that it was 
 not so originally can be seen at once from a comparison 
 between the elaborate ornament on the south door and the 
 plain work of the northern one ; the position of the manor- 
 house points the same way. 
 
 1 Cf. the north and south doorways at Cowley, both Norman. 
 
FIG. 77. THIRTEENTH-CENTURY DOORWAY, MILTON (c. 1240) 
 
190 THE GRAMMAR OF ARCHITECTURE 
 
 The west doorway was the great ceremonial entrance, 
 only opened for processions or for the admission of 
 dignitaries of Church and State. 
 
 Besides the three entrances in the body of the church 
 there was usually a small doorway in the south wall of the 
 chancel for the private use of the priests ; the priest's 
 
 FIG. 78. EARLY FOURTEENTH-CENTURY DOORWAY, DORCHESTER 
 
 door in S.Thomas's Church (Fig. 51) is very interesting 
 because it retains its original ironwork of the thirteenth 
 century. 1 
 
 The Norman doorways were the chief glory of 
 twelfth-century architecture. Their recessed orders, 
 
 1 The ironwork on the door of Merton Hall is also a fine 
 example of a fourteenth-century smith's craftsmanship. 
 
FIG. 79. TOWER OF T^E FIVE ORDERS 
 
192 THE GRAMMAR OF ARCHITECTURE 
 
 originally invented under the stress of rude material, and 
 ruder appliances, were developed into such ornamental 
 features that the later builders, as at Iffley (Figs, n and 
 76) actually thickened the lower part of the wall so as to 
 get in an extra order. Two types of Norman doorways 
 
 80. 
 
 FIFTEENTH-CENTURY DOORWAY, MERTON COLLEGE 
 CHAPEL, A.D. 1424 
 
 may be distinguished, and both are illustrated at Iffley : 
 in one (Fig. 1 1) the orders are continued round the arch 
 and the sides of the opening ; in the other (Fig. 76) the 
 orders of the arch are stopped upon shafts in the jambs. 
 The Norman doorway of S. Peter's Church is of the first 
 
THE DOORWAY 193 
 
 type, and that of the Chapter House shows a combina- 
 tion of the two, the inner orders being continuous from 
 the arch to the floor, the outer ones being provided with 
 shafts. Of the ornaments carved on the orders and capitals 
 we shall treat in a separate chapter. 
 
 In the thinner walls of Gothic buildings deeply recessed 
 doorways became impossible. The thirteenth-century 
 builders did, indeed, contrive to give the effect of deep 
 recessing by setting shafts in jambs widely splayed for the 
 purpose, as may be seen in a fine doorway at Milton (Fig. 77); 
 the contemporary builders in France built the marvellous 
 portals of Amiens and Rheims by thickening the wall for 
 the sake of increasing the number of orders ; but the 
 English Gothic builders (though their doorways are always 
 beautiful) wisely reduced the number now that their walls 
 were thin, and contented themselves with one or two 
 shafts, or three at the most. These are not detached 
 from the jambs except when they are formed of Purbeck 
 marble, as in the west doorway of S. Helen's Church at 
 Abingdon. 
 
 The south entrance to S. Giles's (c. 1200) is our best 
 example of an Early Gothic doorway, though it is not to be 
 compared with the contemporary doorways at Milton and 
 Haseley. We have no illustration of a fourteenth-century 
 doorway, but Sir Gilbert Scott's entrance to the north 
 aisle of S. Mary Magdalene is a good copy of one. 
 
 There are two beautiful little fourteenth-century 
 doorways with shafts in the south aisle at Dorchester 
 (Fig. 78), but after the Early English period recessed 
 doorways became less common, except in the grander 
 buildings. The capitals of fifteenth-century shafts usually 
 have mouldings instead of foliage ; the doorways of 
 Merton Chapel (Fig. 80) and the Divinity School (c. 1450) 
 
 785 N 
 
FIG. 81. QUEEN'S COLLEGE GATEWAY (c. 1710 
 
THE DOORWAY 195 
 
 are among the finest examples of the century. In the 
 Perpendicular period the door-arch is almost always set 
 in a square frame, and the triangular spaces in the corners 
 between the arch and the label are filled with sunk panels 
 or shields of arms. This is the form of the great gateways 
 in all the older college buildings ; those of All Souls 
 (c. 1440), S. John's (c. 1437), and Brasenose (c. 1509), 
 and the west doorway of Magdalen Chapel are typical. 
 
 This form of doorway, minus all its ornaments, and 
 with its arch so depressed that it was almost minus 
 spandrils, remained in fashion until nearly the end of the 
 seventeenth century. The type of doorway seen in Kettel 
 Hall, for instance, or in the Old Schools quadrangle 
 (Fig. 79) was familiar to Elizabeth and all the Stuarts in 
 every manor-house they visited. 
 
 In Classic buildings large doorways were built with 
 semicircular heads, like the Gateway of Queen's College 
 or Wyatt's Gateway, Canterbury Quadrangle ; smaller 
 doorways are usually lintelled ; in neither case was there 
 any splay or recessing of the jambs, the openings, as in 
 Roman work, being cut straight through the wall. 
 
 CHAPTER IX 
 
 THE WINDOW 
 
 THERE are two ways of approaching the study of a 
 language. By the ' direct ' method, now coming into 
 favour, the student first acquires a vocabulary and a stock 
 of common phrases, and then, when he has learned to speak 
 the new language, proceeds to study its structure, and to 
 discover the reason for the changes in the form of the 
 N 2 
 
196 THE GRAMMAR. OF ARCHITECTURE 
 
 same word in different phrases ; e. g. he is taught that it 
 is right to say 'Magister puerum docet', but ' Pueri a 
 magistris docentur ' before he learns anything of declen- 
 sions and conjugations ; that is, he learns the foreign 
 language as he learned his mother tongue, empirically. 
 
 By the older method the pupil began his study with 
 the laws of the structure of the language, with the con- 
 jugation of doceo, and the declension of magister. 
 
 This is the scientific method ; but science is a late 
 growth in human intellect, the empirical method is now 
 considered the more suitable to infant minds, and the 
 scientific to those more mature. 
 
 Until quite lately the study of architecture proceeded 
 upon a system analogous to the ' direct ' method in the 
 teaching of language ; the student first learned to 
 recognize ' words ' and ' phrases ' ; and naturally the 
 most obvious and everyday words were those to which 
 his attention was earliest directed. So the window, the 
 most conspicuous feature in an ancient building, was the 
 feature with which he made the acquaintance of architec- 
 ture, and in all his subsequent study, it was to the windows 
 that he would first turn his attention in visiting a new 
 building. 
 
 Unfortunately for the method the student, as a rule, 
 never got beyond the vocabulary stage, never proceeded 
 to the study of relationships, to what we have called the 
 grammar of architecture. 1 Most students of history can 
 give the approximate date of a mediaeval door or window, 
 but how many can give the etymology of its form ? can 
 trace its relationships to building science as a whole ? 
 A study in which there is no progress, which provides 
 
 1 There is the same danger besetting the new language-teaching : 
 but that is by the way. 
 
THE WINDOW 197 
 
 no exercise for the reasoning powers, is mere dilettantism, 
 of no more value than the smattering of French phrases 
 with which the ' higher grade school ' child rejoices the 
 heart of his innocent parent. 
 
 But I trust that it will have been clear to the reader 
 from the first that the scientific method is the true 
 approach to architecture, and that he will not be surprised 
 to find a chapter on windows at the end instead of at 
 the beginning of this section of the book. 
 
 The window, though a very conspicuous and beautiful 
 feature, is not one of the fundamentals of architecture ; 
 it is a comparatively modern accessory and convenience. 
 There were, and there are, no windows in primitive 
 buildings, in the beehive hut of the Zulu, or the conical 
 wigwam of the Indian ; there were none in the temples 
 of Egypt and Greece ; even in the buildings of luxurious 
 Rome, more light entered from the doorway than from 
 any other source, and open skylights were far more 
 common than windows. 
 
 The window (wind-eye), as the first part of its name 
 implies, seems to have had its origin in apertures pierced 
 in the wall for the sake of ventilation ; such are found in 
 the ruined temples of Egypt and Greece. In Roman 
 buildings they were so placed as to let in a certain amount 
 of light as well as air ; in the early Romanesque churches 
 they were still unglazed, and admitted much more air 
 than light ; and it was not until Norman was passing into 
 Gothic that the second syllable of the word became first 
 in significance. 
 
 Windows may therefore be said to be a Gothic develop- 
 ment ; they are one of the ideas inherited by the Gothic 
 architects and developed by them from germ to perfect 
 organism. 
 
198 THE GRAMMAR OF ARCHITECTURE 
 
 In discussing Saxon windows and, indeed, Saxon work 
 in general, one needs to walk very warily on ground that 
 is yet far from firm. It would be expected a 'priori, that 
 these early windows would be very small, and most of 
 them are so, being often cut through a single stone ; 
 even when the whole window is not so formed the round 
 head is simply a semicircle cut out of one stone, and is 
 therefore not a true arch ; but the heads of many Early 
 Norman windows, and even of small lancets, were often 
 formed in this way, so that the method is no evidence 
 of pre-Conquest work. Again, not a few windows which 
 are generally supposed to belong to Saxon times are fairly 
 large, like those in S. Michael's tower (Fig. 10). 
 
 However, it is no part of our present duty to go into 
 a partially explored country ; the reader may be left to 
 do so on his own account. I will keep within safe limits 
 by saying that some of the windows in Saxon buildings 
 had the characteristics of those in S. Michael's tower. 
 
 Two types are there represented, belfry windows and 
 others, in which some semi-translucent material could be 
 inserted to take at least the edge off the wind. The 
 first type forms one of the few unmistakable features 
 that may always be assigned to Saxon craftsmen. The 
 opening is not recessed, but is cut square through the wall 
 in the Roman manner ; the difficulties of so constructing 
 a large arch are overcome by substituting two small ones, 
 which meet in the middle on a common impost formed 
 by a long stone that runs right through the wall, and is 
 carried by a stone post called a baluster shaft. This very 
 characteristic shaft was made by turning in a lathe, 
 exactly as the leg of a table is turned indeed, its resem- 
 blance to a stout table-leg is very marked. 
 
 We have seen that the greater number of Saxon 
 
THE WINDOW 199 
 
 churches were rebuilt after the Conquest ; no doubt the 
 old material was used again, but there seems to have been 
 so little carved stonework in the buildings that there is 
 usually nothing by which it can be identified as having 
 formed part of an earlier church. The one exception 
 is the turned shaft, and these were often used again 
 by the Norman builders. One such shaft may be seen 
 in a clear-story window in the south transept of the 
 Cathedral. 
 
 The lower windows of S. Michael's tower probably 
 represent the type that lighted the original nave and 
 chancel. The round arches of their heads are clumsily 
 formed of rubble ; the openings contract in the middle 
 of the wall and widen out towards both its exterior and 
 interior faces, so that, whether looked at from within or 
 from without, the windows present a 'splayed ' or funnel- 
 like opening. Glass was not unknown in Saxon times, 
 but it must have been extremely rare in parish churches. 
 A wooden frame on which parchment was stretched, or 
 even a wooden or wicker lattice, was probably the usual 
 light-transmitting medium ; this was placed in the medial 
 opening, where it would be protected by the deep splay 
 of the jambs. 
 
 One other type of window-opening, though not 
 represented in S. Michael's, was common in Saxon 
 buildings. The head was gable-shaped, formed by two 
 long flat stones which met at an angle. Small windows 
 of this sort may often be seen in the walls of barns. In 
 the nave of Bicester Church there is a large triangular- 
 headed opening in the north wall which seems to have 
 been the doorway of a Saxon church. The tower of 
 Caversfield Church near Bicester is certainly Saxon, with 
 very characteristic windows. 
 
200 THE GRAMMAR OF ARCHITECTURE 
 
 Norman windows are better constructed than any of 
 those in S. Michael's tower, their arches and jambs being 
 formed of squared stones. The glass is set near the outer 
 
 FIG. 82. NORMAN WINDOW, SANDFORD (<:. noo) 
 
 face of the wall, and here the opening is very small, for 
 glass was dear and times unsettled. A wide interior splay 
 through the thick wall allows the light to spread inwards. 
 At first the windows were without ornament both within 
 
THE WINDOW 
 
 201 
 
 and without, but in an early window (Fig. 82) at Sandford 
 (c. noo) rude shafts are set in the jambs to carry the outer 
 order of the arch ; twelfth-century windows were often 
 enriched in this way ; those in the west front at Iffley 
 (Fig. n) have shafts in the jambs outside, while those in 
 S.Peter's Church (Figs. 31 and 66) have shafts both within 
 
 FIG. 83. LATE NORMAN WINDOW, CATHEDRAL (c. 1186) 
 
 and without, some of them elaborately carved. In the 
 windows of the north and south walls of Iffley and the 
 chancel at Cassington the recessed orders, like those of 
 the Iffley west doorway, have no shafts, but are enriched 
 with carved chevrons or a bold semicircular moulding. 
 A comparison of the west windows of Iffley with those 
 on the north and south will show that plain windows 
 were often built in Late Norman work, and are therefore 
 no proof of date ; Cassington windows are very plain, 
 
202 THE GRAMMAR OF ARCHITECTURE 
 
 but the fact that the jambs are moulded proves that they 
 are late in the style. Late Norman windows are usually 
 long and narrow, approaching the lancet type in their 
 proportions. 
 
 In the latter half of the twelfth century circular 
 windows were occasionally constructed in the western 
 gables of churches (Fig. n) ; that at Iffley is a modern 
 
 FIG. 84. LATE NORMAN WINDOW, BUTTON (c. 1 1 70) 
 
 restoration, but is believed to be faithful to the original 
 design. The east end of the Cathedral chancel is also 
 a restoration (Fig. 13), but is much more conjectural. 
 However, its wheel window with its radiating members 
 will serve as an illustration of a beautiful type that came 
 into existence in the reign of Henry 11. 
 
 In the arcade of interlacing Norman arches in the south 
 wall of S. Peter's Church (Fig. 66) lancets are formed 
 at each point of intersection. In the tower at Sutton 
 Courtney (Fig. 84) lancets so produced are pierced and 
 
THE WINDOW 
 
 203 
 
 glazed so as to form what are probably the earliest lancet- 
 windows in England. 
 
 Exterior Interior 
 
 FIG. 85. LANCET WINDOW (c. 1220) 
 
 By the end of the twelfth century the lancet type had 
 become everywhere common, and singly or in combinations 
 it prevailed for half a century ; indeed, it may be said to 
 
204 THE GRAMMAR OF ARCHITECTURE 
 
 have dominated the form of the window all through the 
 Middle Ages, for its long narrow proportions were main- 
 tained in the window-lights whatever might be the shape 
 of the composition as a whole. 
 
 The range of windows in the south aisle of S. Giles's 
 Church (c. 1200) affords an excellent illustration of the 
 early lancet, long, narrow, plain both within and without, 
 its glass set near the outer face of the wall, and inside, 
 a wide splay to allow the light to spread. Other good 
 examples may be seen in the Lady Chapel of S. Peter's, 
 at Elsfield, and Wood Eaton. 
 
 The Norman practice of supporting the arch upon 
 shafts in the jambs was continued, to the greater beauty 
 of the window, as may be seen by comparing the plain 
 early lancets of the south aisle of S. Giles's with those 
 on the north side, which are twenty years later (Fig. 14). 
 There, too, will be seen a further development in the 
 arrangement of windows : the lancets are set in groups; 
 outside they appear as separate windows, but inside the 
 jambs are so sloped that a pair or a triplet of lancets forms 
 one composition. Yet even outside there is evidence of a 
 new grouping principle which was destined to produce most 
 wonderful results : the eastern lancets are completely 
 independent ; but those on the north, though each is 
 distinct, are arranged in groups of two or three, each 
 group being covered by a hood-mould common to the 
 members of the group. 
 
 By far the finest group of lancets in Oxford is that in 
 the east window of the Chapter House ; 1 1 use the singular 
 number because most people see the Chapter House from 
 
 1 The effect is rather spoiled by the transoms inserted in the 
 seventeenth century for the purpose of making casement-openings 
 in the lower part of the lights. 
 
THE WINDOW 205 
 
 within, where the group appears as a single composition ; 
 but if the east end is seen from the Canons' Gardens, 
 each of the five lancets appears as a separate window ; 
 close inspection, however, reveals even externally a sug- 
 gestion of unity in a relieving arch built into the wall 
 ^bove the heads of the lights. 
 
 Every detail of the beautiful composition within will 
 repay the minutest study. The mere ornaments will be 
 examined later on, but the general scheme by which 
 five windows are unified into a single whole calls for 
 immediate notice. It will be seen that the end is accom- 
 plished by means of shafts ; these are not placed in the 
 jambs, but support an inner range of arches in the same 
 plane with the inside face of the wall ; it is their unity 
 which gives unity to the whole window ; they give it 
 besides such grace and lightness that the gap in the 
 massive wall becomes a prism of illuminated air. 
 
 This beautiful arrangement marks most of the best 
 lancet-work ; it may be seen in the windows of S. Giles's 
 north aisle, and in those of the Cathedral spire ; it was 
 never completely abandoned by Gothic architects even 
 when they had discovered a new method of ' composing ' 
 a window, but it ceased to be common after the reign 
 of Henry III. 
 
 The next step forward, which was taken towards the 
 middle of the thirteenth century, seems to have been 
 anticipated by the builders of S. Giles's tower. Even 
 before the building of the north aisle, the tower-makers 
 had grouped two lancets together under one arched hood- 
 mould ; but they had been annoyed by the blank space 
 left between the heads of the lancets and the over-arch, 
 and so they pierced it with another small lancet. In 
 this they proved themselves pioneers ; the aisle-builders 
 
206 THE GRAMMAR OF ARCHITECTURE 
 
 twenty years later accepted their suggestion, and in one 
 of their lancet groups they, too, pierced the space between 
 the hood-mould and its pair of lancets. But they improved 
 upon their model, and instead of a third lancet they made 
 
 FIG. 86. THE GERM OF TRACERY, S. GILES'S (c. 1200) 
 
 a small circular opening. Oxford University was then 
 in its early years ; had it numbered among its members 
 a sufficiently able theorist, he might have foretold from 
 this north-western window of S. Giles's the form of the 
 east window of Merton Chapel and have lived to see 
 
THE WINDOW 
 
 207 
 
 his prophecy fulfilled. For this window and similar ones 
 in the spire and belfry of the Cathedral tower (Fig. 27) 
 mark the conception of a new idea in architecture, the 
 idea of tracery in the window-openings. From Late 
 
 FIG. 87. EARLY TRACERY, WOODSTOCK (c. 1240) 
 
 Norman times the builders had been feeling after beauty 
 as well as light in the window, and now they had dis- 
 covered a new principle of beauty. The Norman architect 
 had beautified the form by which the light entered: 
 the Early Gothic builder had improved upon that form. 
 
208 THE GRAMMAR OF ARCHITECTURE 
 
 Now, in the second quarter of the thirteenth century, 
 light was let in, not by one beautiful form of opening, 
 but by a group of such. The history of windows for the 
 next hundred years is the history of the invention of 
 
 FIG. 88. EARLY TRACERY, S. GILES'S (c. 1265) 
 
 beautiful forms which could be combined in an opening 
 to admit light. 
 
 The first experiments were made by piercing the space 
 above the heads of a pair of lancets ; such compositions 
 mark the work of the first half of the century. The simple 
 circle was soon improved upon, and a quatrefoil took its 
 place, as in the windows of the Cathedral spire. Other 
 
THE WINDOW 209 
 
 geometrical forms were also used : there are interesting 
 examples in the towers of Harwell and Brize Norton. 
 Then the builder began to experiment in the space 
 between three lancets and their hood-mould. 
 
 Here a single piercing was not enough ; the space 
 must be filled by a combination of apertures. We 
 find the solution of the problem in the east window of 
 the Lady Chapel of S. Giles's (c. 1265) ; three circular 
 openings are pierced in the triangle between the heads of 
 the lancets and the arch above. I say ' pierced ', but as 
 they so completely fill the space that there is no longer 
 any suggestion of walling in the head of the window, it 
 would be more true to say that in the void above the 
 lights three rings of stone have been inserted. And that 
 is most probably the method by which the window was 
 formed ; a large opening was cut, and three lancets 
 with the circles above them were built up inside it of 
 bars of stone. That certainly is the method adopted in 
 all windows after the middle of the thirteenth century. 
 It is the converse of the lancet system of window com- 
 position : there separate apertures are grouped to form 
 one window; here a single aperture is divided into 
 variously shaped parts by means of a framework of stone 
 bars fitted into it ; if we may apply to concrete processes 
 the terms descriptive of mental operations we might say 
 that the one method was synthetic, the other analytic ; 
 in the progress of window-making, as in that of science, 
 that is the natural order of things. 
 
 TRACERY 
 
 The combinations of openings in the head of a window 
 form what is known as Tracery. The early forms, 
 
 785 O . 
 
210 THE GRAMMAR OF ARCHITECTURE 
 
 produced by piercing the wall, have been called Plate- 
 tracery ; pierced tracery would perhaps be more expres- 
 sive. The patterns formed by curved bars are usually 
 
 FIG. 89. GEOMETRICAL TRACERY, MERTON COLLEGE CHAPEL 
 (c. 1297) 
 
 termed Bar-tracery ; one might suggest inserted tracery 
 as an alternative phrase. 
 
 Plate-tracery, as we have seen, lasted but for a generation 
 in England, and was then improved out of existence. 
 
 The French, however, continued to develop perforated 
 
maam 
 
 FIG. 90. VERY EARLY BAR-TRACERY, NORTHMOOR CHURCH 
 (c. 1270) 
 
212 THE GRAMMAR OF ARCHITECTURE 
 
 tracery long after the English had left it behind in their 
 progress towards the perfect window. Some idea of their 
 late plate-tracery may be gathered from the windows in 
 the modern church of S. Frideswide, and from those 
 in the Congregational Chapel on the Cowley Road. 
 
 Up to this point the development of architecture had 
 gone on part passu in the countries of Western Europe 
 (Fig. 25), but now, as the nations found their individuality, 
 the national styles diverged. 
 
 In the early bar-tracery of the second half of the 
 thirteenth century the simple geometrical forms of the 
 older type were retained, the arches of the windows being 
 filled with trefoils, quatrefoils, and curvilinear triangles of 
 varying proportions and combinations. 1 The plain heads 
 of the lights were also enriched by the insertion of two 
 curved projections called cusps, which gave a trefoil form 
 to the top of the lancet, and made it more harmonious 
 with the tracery above. The early cusps were separate 
 ornaments fitted into the under-side (or soffit) of the 
 arch, like those in the heads of the lights in the east 
 windows of S. Giles's Chapel and the chancel at Hampton 
 Poyle ; but towards the end of the century the arches 
 of the lancets were made in the form of a trefoil and the 
 cusps ceased to be independent ; after the end of the 
 Lancet period the cusps increased in number, so that 
 the head of the light became a cinquefoil. The late lancets 
 in the chancels of Stanton S. John and Piddington are 
 good illustrations of the improved method ; and the one 
 old window in the tower of Carfax shows, by its trefoiled 
 head, that it too belongs to this period. 
 
 1 This is the ' geometrical ' tracery of Sharpe, who seems to have 
 supposed that the more complicated 'curvilinear' designs were 
 drawn free-hand. 
 
THE WINDOW 
 
 213 
 
 The windows of Merton Choir, built at the end of the 
 thirteenth century, are among the most beautiful examples 
 in existence of geometrical tracery ; the lower window 
 
 ' 
 
 FIG. 91. WINDOW, S. MICHAEL'S (c. 1260) 
 
 in the northern face of S. Mary's tower is hardly less 
 beautiful ; and, if the eye is not satisfied with seeing, 
 one can go to Haseley for further gratification, or to 
 Lewknor and Chinnor, which are only a little farther on. 
 
214 THE GRAMMAR OF ARCHITECTURE 
 
 Or one can take a shorter journey and visit the north 
 aisle of S. Mary Magdalene, where Sir Gilbert Scott has 
 reproduced the tracery of those churches. Some people 
 may similarly spare themselves a journey to Northmoor 
 by studying the clear-story windows of the Cowley 
 Fathers' Church, where a most interesting form of early 
 tracery (Fig. 75) has been copied by Mr. Bodley. The 
 windows of the north choir-aisle at Dorchester are also 
 very early examples of the new development. 
 
 But not all windows of the age of Wallace and Bruce 
 were decorated with tracery. The single lancet with 
 trefoiled head still held its own, as at Stanton S. John ; 
 and many windows were simply divided into three lights 
 by plain mullions without tracery ; the east window of 
 S. Giles's and the south window of S. Michael's (Fig. 91) 
 are examples of a composition very common in parish 
 churches in the days of Edward I. A very similar window 
 may be seen in the upper story of S. Mary's tower 
 (c. 1280) ; but here the mullions interlace in the head of 
 the window. When the heads of the lights and the spaces 
 above them are foliated with cusps, as at Dorchester, this 
 simple design produces a very beautiful window. 
 
 CHAPTER X 
 
 THE WINDOW (CONTINUED) 
 
 BEAUTIFUL as are the windows of Merton the vigorous 
 ' Gothic ' mind was not satisfied with one solution of the 
 problem ; the geometric method of producing a perfect 
 window was too easy, it needed but to group geometrical 
 figures in various combinations and to select the most 
 beautiful that presented itself. So now the artist-builder 
 
THE WINDOW 215 
 
 began to conceive new designs not suggested by the simpler 
 geometrical forms, and to plan intricate and sinuous 
 curves by combining parts of many circles. Thus, in 
 the early years of the fourteenth century, a new form of 
 tracery beautified the windows, formed of free-flowing 
 curves, and hence known as the ' flowing ' or ' curvi- 
 linear ' type ; the latter term, invented by Sharpe, has 
 become conventional: but, of course, geometrical tracery 
 is formed by curved lines, and curvilinear tracery by 
 .geometrical curves. 
 
 The invention of flowing tracery completed the unifi- 
 cation of the parts of the window ; in the geometrical 
 period the lights and the traceried head had been distinct 
 portions ; now the mullions flowed into the tracery, and 
 the ogee curves in the heads of the lights were continued 
 in those of the design above. It was the discovery (or 
 rather the re-discovery) of the ogee curve that brought 
 the new tracery into being. 
 
 Our Oxford examples in the Latin Chapel and the 
 north aisle of S. Peter's are not among the earlier or 
 finer designs to be found in England ; they probably 
 belong to a date when the war with France was beginning 
 to fill the minds of Englishmen with ideas of plunder 
 and self-aggrandizement ill-suited to the progress of art. 
 The most beautiful flowing tracery belongs to the reign 
 of Edward II, when it and geometrical tracery were used 
 side by side, as at Milton. Witney, Chipping Norton, 
 Broughton, Adderbury, and Bloxham, have finer examples 
 than any in Oxford itself. 1 
 
 The beauty of a flower is greatest on the day before 
 it withers. The new method, at first invented to produce 
 
 1 The most beautiful windows in England are the east windows 
 of Carlisle and Selby and the west window at York. 
 
216 THE GRAMMAR OF ARCHITECTURE 
 
 more intricate and subtle curves in the window-tracery, 
 was soon seized upon as a means of producing a design 
 without the labour of conceiving it. A very graceful 
 
 FIG. 92. S. MICHAEL'S (c. 1320) 
 
 undulatory curve marks all the flowing tracery of the 
 fourteenth century, and it had been used to produce 
 a design for a two-light window (Fig. 92) that was the 
 most beautiful small window of the Middle Ages ; it is 
 the form of the south-west window in S. Peter's. This 
 
THE WINDOW 
 
 217 
 
 design was now taken as a unit for a large window, and was 
 simply repeated to form a pattern ; so was produced the 
 type of window (Fig. 93) seen in the vestry of Merton 
 
 FIG. 93. S. MARY MAGDALENE (c. 1337) 
 
 (c. 1310) and the south aisles of S. Mary Magdalene and 
 S. Aldate's ; the tracery is a mere network formed by the 
 repetition of a single form, and is therefore known as 
 reticulated tracery. It marks the beginning of the end ; 
 it is the first symptom of paralysis of the Gothic mind. 
 
2i8 THE GRAMMAR OF ARCHITECTURE 
 
 What, then, shall we say of the petrified cabbage-net that 
 does duty for an east window in the modern church of 
 S. Peter-le-Bailey ? 
 
 We saw that tracery had its origin in the desire to 
 beautify the forms of the openings by which the light 
 was admitted ; in flowing tracery this idea had been 
 insensibly abandoned ; it was no longer the openings on 
 which the mind of the artist was centred, but the curve 
 and flow of the bars themselves. By the middle of the 
 century all regard for the form of the openings had been 
 lost, and we have tracery like that in the west window of 
 S. Mary Magdalene, in which the bars twist and writhe 
 in the sinuous curves of leaping flames, while the openings 
 between them are mere formless voids. 
 
 This was a grave and significant error. It marks the 
 breaking away from the great Gothic principle that the 
 artistic design should express the purpose of the construc- 
 tion ; it means the sacrifice of use to a false idea of beauty. 
 The structural functions of tracery are to furnish openings 
 which may be glazed to admit light, to provide a frame 
 for the glass, and to give support to the arch of the window : 
 it is at once obvious from an inspection of the example 
 mentioned that the new ' flamboyant ' type of tracery 
 fulfilled satisfactorily none of these purposes. Its reign 
 was, therefore, very brief, not because the architect 
 repented him of his error, but because he was ceasing to 
 be the master craftsman ; his supremacy over mediaeval 
 art was being challenged by another artist, the glass- 
 maker ; and soon he was to be made servant where he had 
 been master, and the form of his work was to be dictated 
 to him by the requirements of the hew industry. 
 
 We have spoken of the walls as a pictorial Bible ; the 
 glass-makers had for centuries been trying to emulate 
 
THE WINDOW 219 
 
 in their windows the paintings of the walls. Even in 
 Norman times some beginning had been made in the great 
 abbeys to manufacture coloured glass. In the thirteenth 
 century rude figures of the saints, pieced together of 
 coloured fragments, were set in the openings of the 
 tracery even in village churches. But they were small and 
 crude, and they obstructed the light. The examples in 
 the lights of Merton Choir windows will be familiar to 
 every one ; they are intensely interesting, and to some 
 people beautiful. The glass in the windows of the Latin 
 Chapel and S. Lucy's Chapel in the Cathedral belongs, 
 like the tracery, to the time of Crecy and Poitiers the 
 date of the Becket window, at least, cannot be earlier than 
 13 Edward III, in which year the king marshalled on his 
 shield the lilies of France in addition to his own leopards ; 
 and since the French arms are borne on a separate shield 
 and not quartered with England, it is probable that the 
 glass is not much later than the year of their assumption. 
 
 The figures in these windows are still small and crude. 
 But the Oxford glass-makers were behind the times. The 
 craftsmen at Gloucester were already composing larger 
 and more life-like figures in fine clear glass, and had, 
 moreover, discovered the secret of making glass of a silvery 
 whiteness to form a background for their pictures. The 
 monks of Gloucester were just then preparing to rebuild 
 their choir with funds provided by the offerings of pil- 
 grims at the shrine of the unfortunate martyr Edward II, 
 and they were naturally anxious to take advantage of the 
 recent discoveries in pictorial glazing in the windows of 
 the new work. 
 
 Now that larger figures could be made, larger windows 
 were possible : so the new windows must be made wider. 
 But their wide arches would need more support from 
 
220 THE GRAMMAR OF ARCHITECTURE 
 
 the tracery : and the weight-carrying bars in the tracery 
 are the upright mullions between the lights ; these must 
 therefore be carried right up to the head of the window 
 to support the arch. Then what became of the curves 
 and circles in the top of the window ? Obviously their 
 day was over. 
 
 So reasoned the architect. But the glass-maker arrived 
 at the same conclusion by a different line of reasoning. 
 He was able and anxious to turn the window into a picture, 
 but for the success of his composition it was essential 
 that the irregularly shaped, variously sized, and diversely 
 disposed openings should be modified to form a framework 
 for the insertion of his figures. Imagine the despair of 
 an artist who was required to fill the radiating sectors 
 of a wheel-window with full-length figures of the apostles. 
 
 Therefore in the new work at Gloucester, begun in 
 1337, was evolved a new type of tracery in which neither 
 the form of the aperture nor the curves of the bars, but 
 the composition of the stained-glass picture was the pre- 
 dominant motif. Great windows, fifteen or twenty feet 
 in width, and twice as high, are divided into rectangular 
 compartments, each forming a frame for one of the figures 
 composing a complete window-picture. In the lower 
 part of the window ten or twelve long panels contain 
 figures almost of life-size ; above them, tier upon tier, 
 are ranges of smaller panels, each containing a propor- 
 tionate painting. I say panels because the openings have 
 ceased to be lights ; the figures, indeed, are set in a back- 
 ground of white glass, and the huge windows efficiently 
 illuminate the building, but one feels them to be glass 
 pictures rather than windows. Between each tier of 
 panels a horizontal bar of stonework crosses the window 
 from jamb to jamb. It is known as a transom, and its 
 
THE WINDOW 
 
 221 
 
 purpose is to give, both in appearance and in reality, 
 coherence and strength to the framework. 
 
 Seen from without the effect of this form of tracery is 
 that of a stone gridiron. As such it forms the gravamen 
 
 FIG. 94. S. PETER'S IN THE EAST (c. 1360) 
 
 of Ruskin's charge of degeneracy in the fifteenth-century 
 architect ; but, indeed, he might as reasonably have 
 arraigned the gridiron itself for not being a cutlet. The 
 truth is that window, as in the modern sense, now 
 meant the glass and not the framework ; tracery, in the 
 
222 THE GRAMMAR OF ARCHITECTURE 
 
 true meaning of the word, may be said to have come to 
 an end in the fourteenth century. Henceforward window- 
 making ceases to be the province of the architect : he has 
 
 FIG. 95. FIFTEENTH-CENTURY WINDOW, S. MARY'S 
 
 simply to provide a frame for the real window-artist. 
 In the windows of New College Chapel (Fig. 16), and in 
 every window filled with good glass, no one notices the 
 form of the stonework : the more it challenges attention, 
 the less it befits its purpose as setting for the picture. 
 
THE WINDOW 223 
 
 It is now that the picture has been reft from the frame 
 that the latter appears as a window ; as a design for 
 light-openings no excuse can be made for it, but it was 
 never intended, and ought not to be judged, as such. 1 
 
 For fifty years and more the old deeply-rooted concep- 
 tion of a window as a beautiful opening caused the 
 English builders to resist the example of Gloucester. The 
 results of the Black Death and of the failure of the war, 
 and the mental paralysis of Church and State caused 
 by the astounding demands of the labourers for a living 
 wage and an uncorrupt priesthood, also checked the 
 progress of architecture for a whole .generation. Yet 
 the trace of the new influence can be seen in short vertical 
 members inserted to strengthen the flowing tracery after 
 the middle of the century ; such indications of the coming 
 change may be observed in one of the windows of 
 S. Peter's north aisle (Fig. 94), in the Becket window of 
 the Lucy Chapel, and in the curious tracery of the two- 
 light window in the south aisle of S. Giles's Church. 
 
 However, when William of Wykeham planned his new 
 college in 1380, his adoption throughout of the rectangu- 
 lar type gave the coup de grace to curvilinear tracery. 
 
 In the next century Wykeham's glass and its con- 
 comitant grille came everywhere into fashion ; hence the 
 term Perpendicular is commonly used to denote the 
 manner of fifteenth-century work. It is a happier term 
 than many of those invented by the Gothic revivalists, 
 since rectangularity is a well-marked characteristic, not 
 only of the windows, but of the wall-panels, the doorways, 
 and most of the ornamental features as well. 
 
 1 Even so; the fourteenth-century tracery in the Sanctuary at 
 Dorchester is far uglier, now that its glass is gone, than the frankly- 
 confessed panels of the fifteenth century. 
 
224 THE GRAMMAR OF ARCHITECTURE 
 
 CHAPTER XI 
 
 THE WINDOW (CONCLUDED) 
 
 SUCH was the enthusiasm for the new glass that small 
 early windows were often pulled out in order that a 
 building might be lightened and beautified by its means. 
 Five Norman windows were thus sacrificed at IfHey, 
 including the circular window now restored to its 
 original form. The shield of Pole impaling the royal 
 arms in the glass of the south-west window suggests that 
 these windows date from the time of John de la Pole, 
 Duke of Suffolk (1463-91), who married Elizabeth of 
 York. 
 
 The depressed arches of the wide fifteenth-century 
 windows gave so little room in the heads that the builders 
 were forced to make the tracery encroach upon the space 
 below ; therefore, in many examples, the windows of the 
 Divinity School for instance (Fig. 17), the large lights 
 do not rise to the level from which the arch springs ; and, 
 since the tracery descends below that level, it is sometimes 
 spoken of as ' dropped ' tracery. 
 
 Though the stonework of the window had suffered, as 
 an artistic conception, yet there remains much to admire 
 in the scientific disposition of the mullions, especially in 
 the earlier work of the fifteenth century. The windows 
 of New College Chapel are good illustrations ; there are 
 sub-arches in the tracery to lessen the thrusts of the 
 great containing arch ; and the thickness of the mullions 
 is nicely graduated according to the weight of the burden 
 they carry ; their form also should be compared with 
 that of earlier mullions ; it will be seen that their diamond- 
 shaped section, narrow and deep, gives a minimum of 
 
THE WINDOW 225 
 
 surface to wind-pressure with a maximum of strength 
 to resist it. 
 
 I shall have failed in my main purpose if the reader has 
 not realized that down to the point at which we have now 
 arrived there was a continuous progress in the construc- 
 tion of every architectural detail. In the twelfth and 
 thirteenth centuries especially, development had been so 
 regular that it is almost possible to distinguish the work 
 of any decade from that of the corresponding period 
 before and after. But after 1400 there is a slowing down ; 
 change there must have been, but it was so gradual as to 
 be almost imperceptible the years of growth were over. 
 There is, for instance, little in the form of a window of 
 1410, by which it can be distinguished from one of 1480. 
 The windows of S. Mary's are a whole hundred years 
 later than those of New College, but there is- so little 
 difference to show for that century that one might be 
 pardoned for thinking them contemporaneous; so the belfry 
 stories of the towers of Magdalen and Merton differ in 
 date by more than half a century, but it would need a very 
 expert critic to declare from inspection which was the 
 earlier. Knowledge of construction alone, after 1400, is 
 insufficient to enable us to do more than assign a building 
 to some date in the fifteenth century. To date it more 
 definitely we have to turn for help to other branches of 
 archaeology, and especially to heraldry, to the study of 
 which all serious students of architecture must eventually 
 come. For heraldry, like architecture itself, is one of 
 the tongues of history, and their stories are often com- 
 plementary the one telling the date of a building, the 
 other the names of the builders. And where, as often 
 in the fifteenth century, the architectural evidence is 
 indefinite, some shield of arms carved in a boss or spandrel, 
 
 785 P , 
 
226 THE GRAMMAR OF ARCHITECTURE 
 
 or emblazoned in a stained-glass window, will fix the date 
 as well as suggest the founder. 
 
 For a century, then, the form of the window changed 
 but little. 
 
 Then came the great earthquake of the Reformation ; 
 and when the black night of confusion and wreckage 
 was over, the dawn rose upon a transformed world. Every 
 prospect was changed; man had to adapt himself to a new 
 mental environment. 
 
 It is usually said of church organization and church 
 architecture alike that they were hopelessly corrupt and 
 degenerate in root and branch, and that nothing short 
 of a cataclysm would have been sufficient to clear the 
 ground for better things. That question cannot be 
 argued here. But whatever may be thought of the 
 Church, it is certain that architecture never recovered 
 from the blow ; religion went out of it ; it ceased to 
 appeal to the spiritual part of man ; and materialism 
 entered into it ; instead of desire for beauty it proclaims 
 the desire for comfort ; and instead of the pride of the 
 artist in his work it reflects the pride of the paymaster 
 in his possessions ; its message is no longer ' rejoice with 
 me ', but ' envy me '. 
 
 As in the Middle Ages there had been but one type of 
 window, the church window, so after the Reformation 
 there was one type, the house window ; it is the dis- 
 tinguishing feature of the ' domestic Gothic ' which is 
 the one architectural product of the following century. 
 We have seen that all through the Gothic period there 
 was a steady improvement in the lighting of churches, 
 which may or may not have been accompanied by corre- 
 sponding development in the outlook of men. But at the 
 date of the Reformation many of the older churches were 
 
FIG. 96. JACOBEAN WiNDoVs, S. JOHN'S COLLEGE 
 
228 THE GRAMMAR OF ARCHITECTURE 
 
 still insufficiently lighted by their original windows ; 
 and now that the Bible and the Service Book had been 
 put into the hands of the congregation, such mediaeval 
 darkness must be dissipated ; hence the insertion of 
 domestic windows in the walls of so many of our old 
 churches ; they were not needed in the churches in 
 Oxford itself, but Marston, Cowley, Noke, Wood Eaton, 
 Hinksey, Binsey, and many other village churches in the 
 near neighbourhood have ' Reformation windows ' in 
 their ancient walls. 
 
 Almost every college possesses a wing or a block of 
 rooms added at this time ; some colleges, like Wadham, 
 Jesus, University, and S. John's, were wholly built or 
 rebuilt in the domestic Gothic style ; the window of 
 that period is therefore by far the commonest Oxford 
 type (Fig. 96). There must be hundreds upon hundreds 
 of examples, all practically identical in the form of their 
 lights. In the small private rooms a two-light window 
 is usual, just two oblong apertures divided by a mullion 
 and framed in a plain square label ; in the large rooms 
 a window is formed by combining six or eight lights 
 in a double row ; this form of window is intermediate 
 between the great Perpendicular windows and the modern 
 glazed trellis. 
 
 No one can see the Garden Front of S. John's or the 
 Fellows' Buildings at Merton without admitting that 
 the Late Gothic builders did evolve a form of window 
 perfectly suited to domestic requirements ; it is when 
 one sees it in a church that one feels that there is some- 
 thing wrong about it. What that is becomes clear in 
 the light of the rather difficult parable of the man who 
 had no wedding garment. I find that the window and 
 the parable help to explain one another. 
 
230 THE GRAMMAR OF ARCHITECTURE 
 
 The Jacobean builders seem to have felt that the dignity 
 of the House of God demanded a nobler form of window 
 than the domestic type, and they made an effort to revive 
 the arched and traceried opening. In the windows of 
 Wadham Chapel they were so successful in imitating 
 older designs that the work can scarcely be distinguished 
 from that of two centuries before. But where they tried 
 to design forms of tracery for themselves, as in the 
 chapels of Lincoln, University, and Oriel, the results 
 compare badly with the earlier work (Fig. 19). 
 
 In the examples given a new form will be noticed 
 in the tracery that of the ellipse. The mediaeval 
 workman never seems to have mastered the difficulties 
 of ellipse construction. It would have helped him to 
 a solution of the vaulting problem, as may be seen by 
 an examination of the sham vault of S. Andrew's ; but 
 though this was realized, it was tried only in one or two 
 buildings, and was soon abandoned in favour of the other 
 solution. 
 
 One delightful window must be mentioned as a pro- 
 duct of Tudor architecture the oriel, a bay window 
 built out upon corbels so as to give a view on all sides. 
 The beautiful oriel in Christ Church Hall and the still 
 finer one at Magdalen are among the finest examples. 
 
 The mullioned window, essentially a Gothic invention, 
 and perhaps its latest structural achievement, was also 
 the last Gothic form to survive. 1 But by the middle 
 of the seventeenth century all the other Gothic forms 
 were extinct, and before the century was out the mullioned 
 window followed them in to the limbo of antiquated 
 things. The last examples in Oxford are the windows 
 
 1 Indeed, it may be said never to have perished, for it was 
 retained in humble buildings even in the eighteenth century. 
 
THE WINDOW 231 
 
 of the Old Ashmolean, built by Wren in 1682. But these, 
 like the strange windows in the belfry of All Saints' Church 
 (Fig. 22), are so exceptional that they must be regarded 
 rather as revivals than survivals, as atavisms produced by 
 the effect of a Gothic environment on the mind of the 
 builder. 
 
 With the rise of the professional architect in the person 
 of Inigo Jones, the Gothic building with Classic details 
 gave way to the purely Classic type. But since the ancients 
 had but few and small openings for light in the walls, 
 and had therefore left no models, the Renaissance archi- 
 tects were forced to invent a form of window in harmony 
 with the Classic styles. This was done by the earliest 
 of them, Palladio and his school of the Italian Renaissance, 
 and the results formed the models of the English architects.. 
 
 There are two types of Palladian windows, one having 
 an arched, the other a lintelled head. Both are so 
 plain that only a short description will be necessary. 
 They are simply square or arched openings cut straight 
 through the wall without any splay or recesses in the 
 jambs. 
 
 In the arched window the keystone in the crown is 
 usually distinguished by carving, and by being made 
 to project beyond the other voussoirs of the arch ; the 
 windows of the Sheldonian Theatre (Fig. 5) are good 
 examples. The square-headed type has sometimes a low 
 stone pediment projecting over it, as in the windows of 
 the Old Ashmolean. The east front of Worcester College, 
 built in the latter half of the eighteenth century, shows 
 a combination of the two types ; a central arched opening 
 is flanked by two square-headed ones ; the result is some- 
 times known as a Venetian window. Those in the end 
 walls of Christ Church Library are fine examples. 
 
232 THE GRAMMAR OF ARCHITECTURE 
 
 The projecting sill is an eighteenth-century addition 
 to the window-opening ; the bottom of the Gothic 
 window was splayed like the jambs. 
 
 The sliding wooden frame or sash came into use late 
 in the seventeenth century ; the mediaeval window, or 
 that part of it made to open ; swung outwards on its 
 hinges like a door. In the eighteenth century many of 
 the casement windows in the older colleges were fitted 
 with sashes ; in some the mullions were cut out entirely ; 
 in others, e.g. at Lincoln and Wadham, the inner half 
 of the mullion was cut away and the sash fitted in behind 
 it ; the latter is a favourite device of the modern builder, 
 except that he builds a sashed window first and then 
 claps a half-mullion in front to ' Gothicize ' it. There is 
 a difference, too, between mutilating an old window for 
 purposes of modern convenience and building a mutilated 
 obsolete form as an original design. 
 
 Modern windows exhibit a chaos of ancient types, the 
 predominant one being the oblong sashed opening of the 
 eighteenth century. But every age, from Early Norman 
 to Late Renaissance, is represented in our modern build- 
 ings. We have Early Norman windows in the new 
 Roman Catholic Church, late ones in S. Clement's, early 
 thirteenth-century lancets in Holy Trinity Church, plate- 
 tracery in S. Frideswide's, early geometrical windows in 
 the new church at Summertown, late geometrical in 
 Mansfield College Chapel, flowing tracery in the windows 
 of the chapel of Manchester College, and Perpendicular 
 in those of the church at New Hinksey. Then we have 
 domestic Gothic windows in the New Schools, and 
 Palladian windows in the front of Hertford College. 
 And lest any ancient type should fail of representation, 
 we have plate-tracery of the French type in the windows 
 
THE WINDOW 233 
 
 of the Congregational Chapel, and Venetian Gothic in 
 those of the Museum. 
 
 But the typical modern window is the square hole in 
 the wall, through which most of us look out on the world. 
 
 CHAPTER XII 
 
 ARCHITECTURAL ORNAMENT 
 
 WITH this chapter we bring to a close the long analysis 
 of mediaeval building construction that forms the 
 principal section of the book. It is possible that some 
 readers will have found the conclusions suggested by the 
 evolutionary method at variance with their preconceived 
 notions of the origins of architectural forms ; they may 
 object that in our search for reason in architecture we 
 have overlooked the symbolical meaning which is com- 
 monly supposed to be hidden in the details of a mediaeval 
 church. The truth is that the symbolism usually read 
 into features of Gothic architecture has no real existence ; 
 it is purely imaginary and sometimes fantastic to the point 
 of absurdity. Symbolism there is, or rather an inward 
 and spiritual meaning beneath the outward forms, but 
 it lies too deep for the dabbler ; conscious symbolism 
 does not exist at all in Gothic construction ; for the forms 
 of the fundamental parts of a building are governed by 
 their inter-relations and not by caprice ; it is because 
 architectural details are so often studied in isolation that 
 misconceptions arise and are perpetuated. It probably 
 never entered the mind of the designer of the Cathedral 
 that the plan of his church was that of the Cross, upon 
 which the world was redeemed ; l to him it was the most 
 1 Moreover, the Cross of Calvary was probably a Tau. 
 
234 THE GRAMMAR OF ARCHITECTURE 
 
 convenient and scientific disposition of the parts of his 
 building. So in his three stories he had no more thought 
 of the number of Persons in the Trinity than had the 
 designer of the triple lancets in the east end of S. Giles's 
 Church ; a single light was not sufficient ; a pair, occur- 
 ring as it did in the aisle walls, would not give adequate 
 dignity to the eastern front ; therefore he composed 
 a triplet of lights. But if his building had been a finer 
 one he would have built five instead of three : what is 
 the suggested symbolism of the number five ? 
 
 Where intentional symbolism exists at all it must be 
 sought for in the carved ornament, in which alone the 
 fancy of the builder had free play. No one that has seen 
 the carved misericords of fifteenth-century stalls can deny 
 that there is, in many of the figures represented, an 
 intended symbolical meaning, more often than not an 
 old familiar Biblical one, as, for example, that of the vine 
 and its branches ; Mr. Bond has suggested that the 
 grotesque figures, such as are sometimes carved on corbels 
 and gargoyles, may sometimes be symbolical perhaps of 
 the divine power to turn evil and ugly things to good and 
 beautiful ends. Many of them, however, merely reflect 
 the spirit of fun in the workman ; often, as at Cumnor, 
 they look very much like caricatures. 
 
 But to assert, for instance, that the carved oak-leaves 
 on S. Frideswide's shrine were intended to suggest her 
 sojourn among the acorn-eating pigs is to give ground 
 for the suspicion that one's acquaintance with fourteenth- 
 century art is limited to the examples in Oxford 
 Cathedral ; the oak-leaf, the most typical leaf of our 
 woods, is represented in every church where naturalistic 
 foliage occurs. If oak-leaves refer to S. Frideswide's 
 history in one place, what do they mean in another ? 
 
236 THE GRAMMAR OF ARCHITECTURE 
 
 What, too, of the bryony and hawthorn on the tomb ? 
 And why are not the pigs represented to put the matter 
 beyond doubt ? There are pigs among the carved foliage 
 in the chancel at Childrey. ' The choice of the foliage ' is 
 not * obviously made for symbolical reasons ' ; the leaf- 
 forms of the maple, the vine, the oak, and the bryony, are 
 the most beautiful of all that grow ; and therefore they 
 were chosen. 
 
 Of the true spiritual meaning hidden in his work, the 
 carver himself was unconscious otherwise it would not 
 be so well worth searching out. It is the unconscious 
 testimony of the leaves that is important : they are the 
 evidence of the character of the ideas in the minds of 
 men in the early fourteenth century. They tell of a 
 growing civilization, of tranquil life, and peaceful habits, 
 of a new interest in external nature, of that leisured, 
 accurate observation which is the beginning of science, 
 and of a love for what is graceful and tender, which is 
 the inspiration of high and noble art. Compare them 
 with the hawks and war-horses at Iffley, compare too 
 the fidelity of observation shown in the carved hawks 
 with that in the crude unrecognizable flower-forms that 
 accompany them ; and then compare the age of Stephen 
 with that of Edward I : therein may be found some 
 real analogies, some profitable field for the exercise of 
 imagination. Then the foliage of S. Frideswide's shrine 
 will suggest, not the barbarian Algar and the swineherds, 
 but the fathers of English poetry and science, Geoffrey 
 Chaucer and Roger Bacon, and knights as tender as they 
 were brave and true. 
 
 It is often said that the form of the globular ornaments 
 in the niches of S. Mary's spire was invented in honour 
 of Eleanor, wife of Edward I ; the only basis for the sugges- 
 
ARCHITECTURAL ORNAMENT 237 
 
 tion is the synchronization between the introduction of 
 the ornament and the death of the queen. But 'post hoc 
 ergo propter hoc will always be sound logic for some 
 people ; for instance, it is usual for guide-books to 
 describe the cross-legged effigies of thirteenth-century 
 knights (of which there are fine examples at Dorchester 
 and Haseley) as the memorials of Crusaders ; some in- 
 genious writers will even declare from the position of 
 the knees the number of Crusades in which the deceased 
 took part. The remarks sometimes made in front of one 
 of these tombs must almost make the occupant chuckle 
 in his stone coffin. The truth is that all thirteenth- 
 century knights were represented with legs flexed and 
 crossed, the right shoulder thrust forward, the hand in 
 the act of drawing, or perhaps sheathing, the sword ; 
 it was an artistic ' pose ' that threw into prominence 
 the massive limbs of the warrior and showed to advantage 
 the contours of his muscles. When in the fourteenth 
 century the flexible chain-mail gave place to plate-armour, 
 a stiffer pose became necessary, and so the effigies of later 
 date lie flat upon their backs ; to cross the feet would 
 have been an artistic blunder. 
 
 Imagination is a good servant but a bad master : there 
 is scope enough for its exercise in the study of mediaeval 
 architecture, but it becomes ridiculous when it lightly 
 reads the notions of a sophisticated age into the systems 
 of a simpler people. That is the besetting sin even of 
 learned historians, but it is fatal to the right compre- 
 hension of the spirit of the past. 
 
 How imagination, controlled by reason and knowledge, 
 can see through the stones of a building into the mind 
 and soul of the builder may be read in the famous chapter 
 ' On the Nature of Gothic ' in the Stones of Venice. 
 
238 THE GRAMMAR OF ARCHITECTURE 
 
 But that is the imagination of the trained reasoner, not 
 the fancy of the irresponsible trifler. Let me illustrate 
 the contrast between a fanciful and a scientific explanation 
 of a structural detail by an example drawn from the study 
 of organic life. Fancy explains the curiously shaped 
 mandibles of the crossbill by the legend that its ancestor 
 perched upon the Sacred Cross and twisted its beak in 
 a vain endeavour to draw out the nails. Science, observing 
 not one fact alone, but drawing its slow conclusion from 
 many patient observations, relates the form of the beak 
 to the food of the bird, and explains that the crossed 
 mandibles are exactly fitted for breaking open the fir- 
 cones, upon the seeds of which the bird depends for food. 
 Which explanation best reveals the infinite mind of Him 
 * who in wisdom made all things ' ? No one, I suppose, 
 would now regard the old legend as more than a poetical 
 fancy ; but my point is that theories just as fanciful still 
 hold the ground in the study of architecture, misleading 
 many and obstructing the light. 
 
 This must be our excuse for halting so long upon the 
 threshold of the chapter. 
 
 Thereare two typesof ornament, representing beauty 
 concrete and abstract. The one consists of forms derived 
 directly or indirectly from actual objects : the other of 
 purely ideal forms. Both types are represented in 
 mediaeval architecture, the first by ornaments copied 
 from, or suggested by, the shapes of real things, the second 
 by the curved surfaces that are cut upon the stonework 
 to modulate the play of light and shade. The alternating 
 convexities and concavities that give contrasting effects 
 of light and shadow are known as mouldings ; the 
 imitative forms, whether conventional or naturalistic, may 
 be distinguished as enrichments. In early work, ornament 
 
ARCHITECTURAL ORNAMENT 239 
 
 of the concrete type naturally predominates ; as civiliza- 
 tion advances, the abstract forms tend more to be 
 preferred. Thus Romanesque architecture, like Assyrian 
 and Egyptian, is mainly ornamented with rude carv- 
 ings ; while in Gothic, as in Greek work, mouldings are 
 numerous and varied. 
 
 Of Saxon ornament we know but little ; probably 
 because there was very little of it, and certainly because 
 still less has survived. Our only important relic of Saxon 
 architecture is the gaunt and gloomy stronghold of 
 S. Michael's, and of its solitary moulding we shall pre- 
 sently speak ; but an attempt has been made to suggest 
 -that the choir of the Cathedral is substantially a pre- 
 Conquest building, and as the principal evidence adduced 
 is derived from the ornaments of the capitals, we will 
 make brief reference to examples of Saxon decorative 
 carving outside our own district. 
 
 In the church porch at Wantage is a fragment of stone 
 that once formed part of a churchyard cross. It is covered 
 with a curious incised pattern of interlacing circles ; 
 similar carving on crosses, known to belong to Saxon 
 times, in the Durham Museum and in other places 
 in the north of England, prove it to be Saxon work. 
 But almost identical patterns on fragments of British 
 pottery and on bronze shields and armlets suggest that 
 this form of ornament was derived from Celtic sources. 
 It thus forms a strong argument for the theory of the 
 survival of a considerable Celtic element in England, 
 especially as its examples are most numerous in those 
 districts that were latest subdued, when the first fury 
 of conquest had abated. 
 
 Now on some of the capitals of the Cathedral piers, 
 and particularly on that of the middle pier on the north 
 
240 THE GRAMMAR OF ARCHITECTURE 
 
 side of the choir, there is a pattern formed by inter- 
 lacing scrolls of foliage. 
 
 But, unfortunately for the theory that the scrolls 
 prove the Saxon origin of the choir, they bear a much 
 closer resemblance to the leaf-scrolls of Classic ornament 
 than to any Saxon or Celtic design. Rude imitations of 
 Classic scrolls are common on Norman capitals, and it 
 would be quite possible to get together a series of examples 
 showing the gradual recovery of the design through the 
 eleventh and twelfth centuries, beginning with the font 
 at Deerhurst, on which both Celtic and Classic influence 
 is represented, and ending with the Cathedral piers. 
 Further, it is possible to trace the pattern in later work, 
 as any one will admit who compares the scrolls on 
 the Cathedral capitals with the ironwork on the early 
 thirteenth-century door of S. Thomas's Church. But if 
 those capitals really reproduced a Celtic design, and this 
 were to be taken as fixing the date of the building, it 
 would prove, not that the work was done in 1004 under 
 the influence of a Norman architect, but that it was 
 executed at least two centuries earlier when Celtic influ- 
 ence in art prevailed. Which is proving too much. 
 
 No one whose eyes are not blinded by his own pre- 
 possessions can fail to see that the caps both of nave and 
 choir are all of one type, however various may be their 
 decorative foliage. They all, both those with scrolls and 
 those with stiff-stalked leaves, have the volute of foliage 
 at their angles that is a certain mark of work done in the 
 last quarter of the twelfth century. 
 
 We have no example of Saxon decorative art in any 
 Oxford building. 
 
 Norman ornament shows a great variety, but most of 
 it belongs to the concrete type, i.e. it consists rather of 
 
ARCHITECTURAL ORNAMENT 241 
 
 enrichments than of mouldings. In the earlier buildings 
 of the half-century following the Conquest there is very 
 little attempt at decoration. The workmen as yet had 
 not the skill, the tools, nor the models. And though the 
 Norman abbot might rebuild his abbey church, the 
 Norman baron had to build a castle for the king or fortify 
 his manor-house, and secure his holding before he had 
 leisure to turn his attention to the church of the village. 
 
 The only graving tool the workmen had was the hatchet 
 with which he squared his masonry ; therefore, all early 
 Norman ornament consists of such forms as could be cut 
 by this means. The sunk-star pattern on the imposts 
 of the chancel arch of Holywell illustrates one of the 
 most common of the early forms of ornament, and the 
 only one represented in Oxford buildings. It may well 
 belong to Robert d'Oilgi's days. 
 
 But it was when the wise policy of the Lion of Justice 
 had established goodwill between the two races that the 
 wonderful carving that is typical of Norman times was 
 executed. A good deal of it is Byzantine in character, 
 and reflects the influence of the Crusades 1 upon western 
 civilization. Constantinople, of old Byzantium, was then 
 the gateway to theEast, and Constantinople was the capital 
 of the Greek Empire, upon which had fallen a double 
 portion of the architectural spirit of Rome. Western 
 Europe had already recovered the constructional principle 
 of the round arch ; now it received from Byzantium those 
 ornamental forms that had been held, as it were, in trust 
 until the western legatees should be ready to use them. 
 
 So Sir Richard or Sir Roger returned from the East 
 with his head full of designs for ornamenting the new 
 
 1 Or rather the reopening of intercourse with the East which 
 followed the First Crusade. 
 
 785 O 
 
242 THE GRAMMAR OF ARCHITECTURE 
 
 village church that he had vowed to build if the saints 
 granted him a safe home-coming. And the village crafts- 
 man rejoiced with him, and especially over the new tool, 
 the chisel, which should make possible the execution of 
 more delicate and complicated designs. 
 
 FIG. 99. CHEVRON ORNAMENT 
 
 What those designs were can be seen best on the door- 
 ways of Iffley, which form practically a complete dictionary 
 of Norman ornament. The most striking features, the 
 signs of the zodiac on the western doorway, the Centaur 
 and the Mermaid on the southern, the leaf-scrolls, and 
 
 FIG. 100. CHEVRON ORNAMENT 
 
 the foliage generally, are Greek in conception, but the 
 treatment and execution are thoroughly native. The 
 beak-heads on the west, found also on the doorways of 
 S. Peter's and S. Ebbe's, are very characteristic of Late 
 Norman ornament. Of their origin little is known. They 
 have been said to symbolize evil spirits waiting at the 
 threshold to pluck away the good seed from the hearts 
 
ARCHITECTURAL ORNAMENT 
 
 243 
 
 of those leaving the building. 1 But this supposition is 
 not very complimentary to the dignitaries of the Church, 
 for whom alone the west doorway was opened. 
 
 The more conventional enrichments may be particu- 
 larized, but the illustrations render unnecessary any 
 
 FIG. 101. NORMAN PELLETS (North doorway, Iffley) 
 
 description. The chevron or zigzag is the most typical 
 ornament of the Norman style ; it was probably derived 
 from Roman work, but its ultimate origin is prehistoric ; 
 it is found on the pottery of neolithic man. It was the 
 last Norman ornament to be abandoned by the Gothic 
 
 FIG. 102. DOG-TOOTH ORNAMENT 
 
 builders, and a very deeply under-cut form of chevron, 
 illustrated in the west doorway of Cuddesdon Church, is 
 one of the marks of the latest transitional stage. 
 
 The Nail-head, a small square pyramid, also survived 
 in Gothic work, but changed and beautified by the 
 chisel into the shape of four leaves, forming what is known 
 1 Matt. xiii. 4 and 19. 
 
 Q 2 
 
244 THE GRAMMAR OF ARCHITECTURE 
 
 as the dog-tooth ornament ; this is a mark of the earliest 
 Gothic (c. 1180-1220), and may be seen on the font of 
 S. Giles's Church, and in the mouldings of the windows 
 of the Chapter House. 
 
 The pellets of the north door of Iffley, the billets of 
 
 FIG. 103. BEAK-HEADS 
 Suggesting a possible origin in the wedge ornament (cf . Fig. 76). 
 
 the southern string-course of S. Peter's (Fig. 66), and 
 the ubiquitous chevron, though they disappeared from 
 architecture after the twelfth century, were retained in 
 heraldry as charges upon shields ; the arms of the Oxford- 
 shire Dormers, for instance, show ten billets, those of the 
 
 FIG. 104. BEAK-HEADS, S. EBBE'S CHURCH 
 
 Giffards as many pellets, and those of Wykeham two 
 chevrons. 
 
 Among the less common enrichments are the chain orna- 
 ment of S. Peter's chancel (Fig. 31), the cable-moulding, 
 and the roses of the south doorway of Iffley, and the 
 wedge ornament (Fig. 76), which may have suggested 
 the form of the beak-head. 
 
ARCHITECTURAL ORNAMENT 245 
 
 It should be understood that, though Norman orna- 
 ments may be classified under the types mentioned, the 
 execution of the form varies with each example. It must 
 be obvious that the carving on the Iffley doorways, for 
 instance, was never done from a detailed drawing, but 
 from the workman's interpretation of a rough sketch, 
 probably even from his own rude design. So it is full 
 of imperfections, like a child's drawing ; but withal it is 
 artists' work, the brain that conceived it moved the hand 
 that shaped it forth. Just as a child's own expression of 
 his idea of a cow is more interesting than his copied 
 
 FIG. 105. THIRTEENTH-CENTURY HOLLOW MOULDING 
 
 drawing of one, so the Norman carving of Iffley and 
 S. Peter's charms by the very originality of its faults. 
 
 Norman mouldings are much less varied than the 
 enrichments of the style ; the earliest, which remained 
 to the end the commonest, merely consists of a square 
 projection, the lower edge of which is chamfered off below 
 a narrow groove or quirk (Fig. 70). This is the abacus- 
 moulding on every Norman capital and impost ; it is 
 cut upon the imposts of the arches in the belfry windows 
 of S. Michael's, and affords clear proof 'that- the work 
 was done under Norman influence, i.e. at the earliest, 
 shortly before the Conquest. 
 
 In late work a bold semicircular bead or torus 1 is the 
 
 1 Also called a bowtell, boutel, or boltell. 
 
246 THE GRAMMAR OF ARCHITECTURE 
 
 commonest moulding ; it is cut upon the outer orders 
 of the arch in the Chapter House doorway, and upon the 
 arches in the choir in the Cathedral. This moulding 
 was retained in the string-courses of plain Early Gothic 
 buildings, but before the end of the twelfth century 
 
 FIG. 106. 
 
 BALL-FLOWER ORNAMENT, PRIOR BUTTON'S TOMB 
 (c. 1300) 
 
 the workmen, . having by then acquired a complete 
 mastery of the new tool (the chisel), had developed from 
 it a pointed bowtell or keel-moulding. This is the 
 typical moulding of the transition period (Fig. 71) ; it 
 is represented in the Cathedral in the arches of the nave. 
 The mouldings of the north doorway in Holton Church, 
 
ARCHITECTURAL ORNAMENT 2 ^ 
 
 of the south doorway at Haseley, and of the tower- 
 arches at Cumnor and Cuddesdon, are excellent examples 
 of this date. 
 
 The thirteenth-century builders proceeded to cut still 
 deeper into the under-side of the bowtell, so producing 
 the deep hollows that distinguish the Early Gothic mould- 
 ings illustrated in the arches of the Lady Chapel and the 
 Chapter House of Christ Church. Mouldings are a much 
 more important feature in Gothic than in Norman work, 
 and enrichments are correspondingly fewer. The dog- 
 
 FIG. 107. FILLETS 
 
 tooth ornament, already referred to, is by far the com- 
 monest enrichment of Early Gothic work ; it is used 
 profusely in the hollows of the mouldings of the Chapter 
 House. It went out of fashion after the middle of the 
 thirteenth century, and its place was taken by the ball- 
 flower ornament (Fig. 106), which is the characteristic 
 enrichment of hollow mouldings in the Edwardian period. 
 It is often said to be derived from the pomegranate, but 
 it bears a much stronger resemblance to the globular 
 bells on the trappings of pilgrims' mules, represented 
 to-day on the martingale of the cab-horse. 
 
 A contemporary but less common enrichment is a four- 
 leaved flower, something like the clematis. It may be seen 
 on the ridge-ribs of the Latin Chapel, in doorways at Dor- 
 chester (Fig. 78) and in a beautiful doorway at Bampton, 
 where it is used in combination with the ball-flower. 
 
248 THE GRAMMAR OF ARCHITECTURE 
 
 The mouldings of the mid-Gothic period are not so 
 deeply cut as those of the Lancet stage, the hollows are 
 wider and shallower, and the convex surfaces seldom 
 show more than a quarter of the curve of a circle ; if 
 a bowtell occurs its face is divided into segments by fillets, 
 little projecting flat-topped ridges very characteristic of 
 the mouldings of the Early English and early Decorated 
 periods (Fig. 107). The commonest moulding of the 
 mid-Gothic stage is known as the scroll (Fig. 73) because its 
 section bears some resemblance to that of a rolled parch- 
 ment cut transversely ; it will be easily recognized in the 
 hood-moulding of the tower-arches in Merton Chapel. The 
 whole of the mouldings on the capitals, bases, and arches 
 of the tower piers are fine examples of their date (c. 1 300) ; 
 the mouldings of the sedilia in the chapel illustrate the 
 beautiful shading that is not the hard black and white 
 of 1250 nor the soft greys of 1350, but is intermediate 
 between them. In the Latin Chapel at Christ Church the 
 scroll-moulding may be seen on the capitals of the vaulting- 
 shafts, and the quarter-round upon the arches. Probably 
 the Latin Chapel is the only building existing in which 
 one can compare completely the mouldings of the Norman, 
 Early Gothic, and mid-Gothic styles without moving 
 from one's seat, or even turning one's head. 
 
 After the middle of the fourteenth century the mould- 
 ings grew wider and shallower ; the ogee curve that had 
 transformed the window-tracery now made its appearance 
 in the mouldings ; a broad waved surface became common 
 and remained so through the fifteenth century. Per- 
 pendicular mouldings almost always show shallow ogee 
 curves ; a very typical form resembles the double ogee 
 formed by the curving pages of an open book before use 
 has made them lie flat. It is illustrated in the mouldings 
 
FIG. 108. CANOPIES OF SEDILIA, MERTON COLLEGE CHAPEL (c. 1300) 
 Showing Decorated foliage and mouldings. 
 
250 THE GRAMMAR OF ARCHITECTURE 
 
 of the north arcade in S . Mary Magdalene Church (Fig. 1 10). 
 But the most easily distinguished mark of fifteenth- 
 century mouldings is a very wide and shallow hollow 
 dividing one suite of mouldings from another. It is 
 
 FIG. 109. OGEE MOULDING (c. 1350) 
 
 particularly noticeable in the window-openings ; the 
 glass is protected by being set further back in the wall 
 than in earlier work, and the jambs are scooped out to 
 form shallow recesses. The windows of S. Mary's Church 
 and of Magdalen Tower afford typical illustrations of 
 this wide hollow moulding (Fig. in) 
 
 FIG. no. DOUBLE OGEE, FIFTEENTH CENTURY 
 
 I have left until last the most beautiful ornament of 
 Gothic architecture, the carved foliage of the capitals, 
 the bosses, and the spandrels. 
 
 There is very little foliage in Norman ornament ; what 
 there is is always conventional, and is imitated from Roman 
 models. But towards the end of the twelfth century, 
 when Norman was passing into Gothic, there is evidence 
 
ARCHITECTURAL ORNAMENT 251 
 
 of a strong desire to return to Nature for models in 
 decoration. It is very noticeable in the foliage of tlie 
 Cathedral capitals ; even in the carving of Classical leaf- 
 scrolls the foliage seems to be struggling to take its 
 own wilful curves, and to break away from the perfect 
 symmetry of Greek leafage ; in the foliage of the later 
 capitals in the nave there is no attempt at symmetry, the 
 leaves have the freedom arid abandon of natural growth. 
 But the builders of the next generation realized that a leaf 
 on a plant is one thing and a leaf applied to architectural 
 ornament is another ; they felt that the foliage on a capital 
 
 FIG. in. FIFTEENTH-CENTURY HOLLOW MOULDING 
 
 should support the abacus as well as decorate the bell, 
 and so they produced that beautiful leafage of the Chapter 
 House and the Lady Chapel which has all the grace and 
 freedom of nature, and yet has constructional meaning 
 too. Theirs is the perfect capital. Their successors who 
 cut the lovely bosses on S. Frideswide's shrine were so 
 obsessed by the beauty of the actual leaf that they copied 
 it in facsimile in their stonework ; it was an amiable 
 weakness ; but it was a weakness. The carvers of stiff- 
 stalked foliage had loved reality too we can see it in the 
 bosses of the Chapter House vaulting but they had 
 self-restraint enough to set construction before ornament. 
 The mid-Gothic artists carved naturalistic foliage on 
 their capitals, disregarding constructional expression in 
 their enthusiasm for botanical accuracy ; and the final 
 
252 THE GRAMMAR OF ARCHITECTURE 
 
 result was that the fifteenth-century builders forgot that 
 foliage had ever been constructional, and having no 
 enthusiasm for natural forms, produced the useless, lifeless 
 carving of the capitals in the Cathedral cloisters. The 
 foliage of the fifteenth century partakes of the square- 
 ness that marks all the other details ; lifelessness follows 
 as a matter of course the right angle is not found in 
 living matter. All the plant-forms of this period are 
 conventionally treated, but in the early work of the 
 century it is usually possible to identify the natural leaf 
 that furnished the base of the design. Thus on the 
 large tomb in the Lady Chapel, sometimes (but errone- 
 ously) spoken of as the Watching Chamber (c. 1424), there 
 is a design from vine-leaves and fruit ; in the carvings 
 of the sedilia in S. Mary's Church (1488) there is a very 
 similar design, but the leaves are unrecognizable. The 
 cresting of the sedilia canopy, and also that of the con- 
 temporary reredos, illustrates a square-leaved ornament 
 that is very common in such positions : it is known as the 
 Tudor-flower, though it is not in the least like a flower, 
 and was in use all through the fifteenth century. It is 
 always used in long suites, and it is possible that it ought 
 to be considered as an ornamental development of the 
 battlement (Fig. 112). 
 
 The setting of the door-arch in a square frame produced 
 triangular spandrels (Fig. 113), which had to be ornamen- 
 tally treated ; shields of arms were often used to fill these 
 spaces, as in many of the college gateways, or a panel 
 was sunk in such spandrel, as in the water-gate of Rewley 
 Abbey ; but sometimes, e. g. in the spandrels of the 
 archways leading to the Old Schools Quadrangle, a large 
 triangular leaf was carved instead. 
 
 The crockets that enrich the outline of Gothic pinnacles 
 
ARCHITECTURAL ORNAMENT 253 
 
 are always derived from foliage. They are not found 
 in Norman work, and are not common in Early Gothic, 
 
 FIG. 112. SEDILIA, S. MARY'S (c. 1488) 
 
 but in the Edwardian period they are used profusely in 
 the canopies of tombs and sedilia, and on the hood- 
 moulds of arches ; there can be few examples as beautiful 
 
254 THE GRAMMAR OF ARCHITECTURE 
 
 as those carved on the canopies of the sedilia in Merton 
 College Chapel. Unlike the other leaf-forms of the mid- 
 
 FIG. 113. SCHOOLS GATEWAY 
 
 Gothic period the crockets are usually conventionally 
 treated. The numerous pinnacles of Magdalen, S. Mary's, 
 
ARCHITECTURAL ORNAMENT 255 
 
 and the Divinity School bristle with crockets, but, like 
 the rest of the leaf-carving of their date, they are very 
 poorly cut. 
 
 The ornaments of the Renaissance period need little 
 notice here, since they are almost all derived from the 
 Greek and Roman Orders already described. 
 
 Gothic mouldings perished with the other ornaments 
 of the style in the sixteenth century. They flickered 
 into life again in Oxford under the influence of the 
 Laudian revival, and those of the gateways of University 
 College compare favourably with work of the fifteenth 
 century. The contemporary gateway of the Schools Quad- 
 rangle (Fig. 113) also shows very inter estingrevivalof Gothic 
 mouldings ; they are so boldly cut that at a first glance 
 they suggest the mouldings of the thirteenth century. 
 But no thirteenth-century builder would have carried 
 them continuously round the doorway ; he would have 
 stopped them upon shafts in the jambs, to the great 
 improvement of the design. 
 
 Mouldings persisted upon the mullions to the last, or 
 rather a moulding, for there was but one type in universal 
 use. The mullion was square in section and had four 
 quarter-rounds cut upon it with wide fillets between them. 
 The Gothic mullion had been formed to resist wind- 
 pressure, and had therefore been made lozenge-shaped 
 in section and with concave faces. But the mullions of 
 seventeenth-century flat-headed windows had to carry 
 a lintel, and hence they were necessarily stouter, and were 
 given a square form with convex mouldings, as in the 
 windows of University and Oriel Colleges and those of 
 the Jacobean houses in Holywell Street, e.g. No. 13. 
 
256 THE GRAMMAR OF ARCHITECTURE 
 
 CHAPTER XIII 
 ARMORY IN ARCHITECTURE 
 
 THE heraldry in our ancient buildings is too big a sub- 
 ject to be discussed at the end of a small book, yet no 
 account of architectural ornament would be complete 
 without some reference to the armorial insignia which 
 formed so large a part of it. From the thirteenth century 
 to the sixteenth, shields of arms, crests, and badges both 
 decorated every building and bore witness to its founders. 
 Our churches are full of blank shields that now, like the 
 empty niches, testify to the fanaticism of the destroyer 
 but preserve no memory of the builder. 
 
 We have already spoken of the pictorial teaching in the 
 painted walls and windows ; scriptural truths formed only 
 one part of the information therein displayed. Above 
 the figures of saints in the window-lights the arms of the 
 givers of the glass glowed in the openings of the tracery ; 
 the stone corbels that supported the timbers of the roof 
 were cut into the form of shields and painted with the 
 coats of the founders, the reigning king or the contem- 
 porary bishop ; the drip-stones of the windows often 
 terminated in similar shields ; and on the bosses of the 
 vaulting they were carved again. Every tomb, of course, 
 showed the arms and crest of its occupant, with those of 
 his wife's family and often of his connexions. 
 
 Inscriptions, though sometimes added, were superfluous, 
 for few could read them, whereas heraldry was a tongue 
 familiar to all. Nowadays the shield of the sovereign is 
 the only one that most people know, but in the Middle 
 Ages, if you knew the name of a man you knew his shield 
 also ; his shield stood for his name, it was his signature, 
 
ARMORY IN ARCHITECTURE 257 
 
 in a sense his sign manual, for it was reproduced upon the 
 seal with which he stamped all documents whether he 
 was a"ble to write or not even, indeed, if he had written 
 them himself. So John, though he could write, did not 
 set his name to the Great Charter ; he signed it with 
 the shield by which he was known to all his subjects. 
 A written signature might be repudiated or forged, but 
 it was almost impossible successfully to forge a seal. 
 
 The instinctive tendency to adopt some private device 
 or symbol as a mark of personal identity is seen in the 
 totems and tatooings of primitive man, mixed up with 
 religious and social notions ; in the tokens ascribed by 
 Homer to his heroes ; and in the distinctive pennons 
 carried by the figures in the Bayeux Tapestry. But it 
 was the development of armour in the last years of the 
 twelfth century that made such ensigns an absolute 
 necessity to the military chief. The great closed helm 
 which then came into use covered the head and rested 
 upon the shoulders ; it therefore became necessary that 
 every leader should assume some distinctive marks by 
 which he might be known to his own men and to the 
 other chiefs with whom he acted. These were displayed 
 in the most conspicuous position, viz. upon the great 
 shield which, until the development of plate-armour 
 in the late fourteenth century, was slung at the breast 
 of every fully-armed gentleman ; they were therefore 
 known as ' armorial bearings ', or shortly, as ' arms ', 
 and the study of their forms, their ownership, and their 
 inter-relations was called ' armory '. 
 
 The military co-operation of Christians of all nations in 
 the Crusades made distinctive arms still more imperative. 
 The scorching sun of Palestine forced the mail-clad 
 knights to cover their steel harness with linen surplices 
 
 785 R 
 
258 THE GRAMMAR OF ARCHITECTURE 
 
 (surcoats) upon which the devices on their shields were 
 reproduced in embroidery. Hence ' shields of arms ' 
 began to be known also as ' coats of arms ', and their 
 devices as ' coat-armour '. 
 
 But the most interesting traces of the influence of the 
 Crusades in armory are to be found in the cognizances 
 adopted by individual Crusaders and borne still upon the 
 shields of their descendants. 
 
 All markings upon a shield are known as ' charges '. 
 Setting aside the primitive totem marks, the earliest 
 charges (though this is still a disputed matter) appear to 
 be derived from the structural features of the shield, the 
 stiffening rim, the transverse, longitudinal, diagonal, or 
 crossed braces by which its form was strengthened, and 
 the brazen studs and bosses by which its bull's-hide face 
 was protected. The first inventors of armorial devices 
 seem to have distinguished their shields by selecting 
 certain of these structural details and accentuating them 
 by defining them in one colour upon the face of a shield 
 painted in another. Thus the ancestor of the Harcourts 
 was known by the two golden bars upon the red field of 
 his shield's surface ; the arms of Balliol College still show 
 the red shield of their founder charged with a rim of 
 silver ; the parti-coloured chevrons of Merton are pre- 
 served in the arms of his college ; the diagonal brace 
 (bend) painted in black upon a silver shield by the ancestor 
 of Radcliffe may be seen in the hall of University College, 
 and the signboard of the ' Osney Arms ' shows two such 
 braces painted in blue upon a gold field, representing the 
 shield of the D'Oilgi's, founders and patrons of Osney 
 Abbey. 
 
 These structurally-derived charges are so common that 
 they are known as the ' ordinaries ' of armory. Whenever 
 
V 
 
 Chief. a. Pale. 3. Pallets. 4. Bend. 5 . Bjndlct 6. Bend sinister. 
 
 8 Bars. 9. Chevron. to. Chevronels. n. Cross. to. Saltire. 
 
 Pile. 14. Quarter. 1 
 
 16. Bordurr. i-j. Qrle. 
 
 FIG. 114. THE ORDINARIES OF ARMORY. 
 
260 THE GRAMMAR OF ARCHITECTURE 
 
 they appear upon a shield with other charges they are 
 always mentioned first in describing it. They were 
 necessarily so few that their combinations were soon 
 exhausted and new charges had to be sought for. These 
 were found by the Crusaders in forms which would serve 
 at once to distinguish their shields and to preserve the 
 memory of their pious enterprise. So the lion, dweller 
 in southern and eastern deserts, appeared in the west and 
 north upon the shields of the kings of England and 
 Scotland ; William the Lion painted a red rampant lion 
 on his golden shield : Cceur de Lion painted two golden 
 lions in the same attitude upon a red shield, but subse- 
 quently increased their number to three and bore them 
 ' passant ' as we now see them. Lions in various numbers, 
 attitudes, and colours were adopted as charges by many 
 of the barons of Western Europe. Lions' heads and lions' 
 paws were selected by others. 
 
 The leathern water-bottles carried by the Crusaders 
 on their marches were also chosen as charges ; there are 
 still several inn signs near Oxford painted with this device. 
 Another favourite charge was the scallop-shell, emblem 
 of S. James, Bishop of Jerusalem, and so of all who made 
 pilgrimage thereto. The shield of Villiers in the entry 
 of the Town Hall shows five golden scallop-shells charged 
 upon a red cross by a crusading ancestor of the Earl of 
 Jersey. 
 
 But almost from the first the choice of charges appears 
 to have been influenced by a desire to make the device 
 on the shield suggest the name of the bearer. Thus at 
 Cumnor the arms of Forster show three bugle-horns, 
 suggesting forester, the original form of the name ; the 
 triple points of the lozenges on the shield of Montagu 
 in the Latin Chapel are a pictorial pun on the older form, 
 
ARMORY IN ARCHITECTURE 261 
 
 Montacute the pointed mountain ; the arms of Egles- 
 field show three eagles ; and the shields of the Butlers, 
 a numerous clan well represented among Oxford worthies, 
 usually bear wine cups as their principal charge. But 
 perhaps no better example of this type of charge can 
 be found than the Ox and Ford on the shield of 
 our city. 
 
 At first each individual consulted only his own fancy 
 in the choice of the device he assumed, but obviously this 
 must have resulted in confusion ' six Richmonds in the 
 field ' would be as embarrassing to friend as to foe. 
 A central authority was therefore necessary to register the 
 devices already assumed, to control the assumption of 
 cognizances by new leaders, and to make uniform rules 
 for what must now become an organized science. The 
 officers to whom this duty was deputed were the 
 heralds of the king and the great officers of the realm. 
 The subject of their jurisdiction therefore came to be 
 known as heraldry. 
 
 Originally the heralds, as the name implies, had simply 
 been messengers employed by kings and nobles in com- 
 municating with friends and foes. But their experience 
 in this capacity had made them specially familiar with 
 armorial cognizances, and so fitted them to act as arbiters 
 in matters connected with the new science. Yet as they 
 were usually uneducated men they fell at first into many 
 absurdities. They knew S. George as a valiant knight 
 who fought under shield, and they supposed that his 
 shield must have borne some cognizance. But as this, 
 unfortunately, had fallen out of recollection they invented 
 it anew, the red cross on the silver field that still forms 
 the principal charge upon the Union Jack. Alfred the 
 Great was a king and a notable warrior (though he did 
 
262 THE GRAMMAR OF ARCHITECTURE 
 
 not fight in a closed helmet) : he must have borne a shield 
 of arms. So this was re-invented for him, a gold cross 
 upon an azure field. And since he is not known to be 
 the founder of University College, that body bears his 
 shield, differenced by the addition of four martlets, unto 
 this day. 
 
 In spite of these and similar aberrations, the early 
 heralds had reduced their system to a science before the 
 end of the thirteenth century. It is no part of my present 
 business to go into the laws and details of that science. 
 My purpose was only to explain the origin of those 
 armorial forms that are so conspicuous in architectural 
 ornament and which, to those who understand them, add 
 so much to the human interest of our old buildings. 
 Incidentally I wished to show that heraldry, so far from 
 being an abstruse and difficult study, is concerned with 
 the symbols of a romantic and picturesque age, was the 
 invention of primitive and simple people, and has no 
 mysteries that cannot be mastered by a very little study 
 of the shields in our college halls and chapels, guided by 
 one of the numerous handbooks available. 
 
 A few words in conclusion may be said of the accom- 
 paniments of the coat of arms that are often confounded 
 with it. The chief of these were the crest, the badge, and 
 the supporters of the shield. Crests existed long before 
 heraldry and, indeed, before man, for they are found on 
 the heads of many birds and of some- animals. There 
 the crest serves the twofold purpose for which it was 
 placed upon the knightly helm : it lends dignity and 
 splendour to the wearer, and it defends the head by 
 attracting blows to itself and so diverting them from it. 
 The comb of the domestic cock is an excellent example, 
 and it is significant that the cock's head, or the head of 
 
ARMORY IN ARCHITECTURE 263 
 
 one of the fabulous creatures derived from that bird, is 
 very common as a knightly crest. The heads of horned 
 animals are also common the crest on the helmet of 
 Sir George Nowers in the Cathedral, for example, is 
 a bull's head. Obviously a crest must always represent 
 some form that might naturally be placed upon the helm. 
 
 Crests were even more strictly personal devices than 
 arms ; a woman or a corporate body might assume a shield 
 and grave its charges upon a seal as a distinguishing 
 cognizance. But a crest is inseparable from a helmet, 
 and a helmet is meaningless except in actual war ; no 
 woman or corporation, therefore, could or can possess 
 a crest though a woman, being the heiress of her father, 
 could transmit his crest to her son. It was reserved for 
 the gentry of our own day to remove the crest from the 
 helmets of their forefathers and place it upon the posterior 
 buttons of their footmen. 
 
 Some mark by which the retainers of a great house 
 might be known was, of course, necessary in the Middle 
 Ages. But a mediaeval lord would no more have used 
 his arms or crest for this purpose than he would have 
 entrusted to another the defence of his own honour. 
 This want was met by the adoption of badges, emblems 
 not so much of persons as of families. Two famous 
 badges are the bear and the ragged staff, used separately 
 or in combination by the great house of Warwick. The 
 inn sign at Cumnor on which they are represented will 
 be known to every reader. 
 
 The fifteenth century was the age of great barons. 
 Magdalen College, founded during the Wars of the Roses 
 (so called, be it remembered, not from any mythical 
 quarrel in a rose-garden, but from the badges worn by 
 the adherents of the rival houses), is rich in examples 
 
264 THE GRAMMAR OF ARCHITECTURE 
 
 of badges used to ornament its buildings. On the bosses 
 of the Founder's Tower may be seen the Shining Sun of 
 York, the Red Rose of Lancaster, and more interesting 
 still, the Rose in the Sun combining the rival emblems. 
 The gable of the buildings facing Magdalen Bridge shows 
 the Portcullis badge inherited by Henry VII from the 
 House of Beaufort. The motto associated with it in- 
 volves the historic pun * altera securitas ' : just as the 
 portcullis formed a second safeguard if the great gates 
 of a castle should be forced, so his mother's descent 
 from Edward III secured the throne to Henry if his 
 own claims should be overborne. In the new build- 
 ings Mr. Bodley has employed several Tudor badges in 
 his scheme of ornament ; among them may be noted the 
 White Hart, Chained, adopted by Richard II, and used 
 by many of his successors. In the spandrels of the doorway 
 of Jesus College Chapel a very interesting combination 
 of badges shows that the building dates from the time of 
 James I ; after his accession to the English throne he used 
 as his badge the Rose of England dimidiated (i.e. halved) 
 with the Thistle of Scotland. Perhaps the most famous 
 badge in English history is that of the Prince of Wales, the 
 three ostrich plumes of the Black Prince, derived, not as 
 in the popular legend from the King of Bohemia, but 
 from his mother's family of Hainault who in turn derived 
 it from the Counts of Ostrevant. 
 
 Badges, like other armorial devices, were sometimes 
 pictorial puns. Thus the badge of Bishop Beckyngton 
 at Lincoln College is a fire-beacon planted in a tun. So, 
 too, the family motto often bore punning reference to 
 their name: ' Do no yll 9 was the motto of D'Oigli 
 (do-illy), ' Ne vile velis ' that of Neville. 
 
 Supporters appeared late in the history of heraldry ; 
 
ARMORY IN ARCHITECTURE 265 
 
 they are figures placed on either side of the representations 
 of shields to suggest protection and display of the arms. 
 Obviously they are the luxuries of peace rather than the 
 necessities of war. The knight carried his own shield 
 in battle, but his squire bore it to the field ; so when he 
 caused its form to be set up in church or hall he gave it 
 added dignity by placing figures at the sides as defenders 
 and exhibitors. These were sometimes men, e.g. the 
 savages that support the shield of Bertie, Earls of Abing- 
 don, sometimes beasts like the Lion and the Unicorn, 
 known to every one as the supporters of the royal shield, 
 and sometimes supernatural agents, as angels, like the 
 shield-bearing figures in S. Mary's Church, or fabulous 
 creatures, like the wyverns that support the shield of 
 Marlborough. 
 
 Supporters are the exclusive privileges of the great. 
 The sovereign, peers of the realm, and knights of the 
 ancient orders are, with a few exceptions, alone entitled 
 to exhibit them. Some great corporations (of which 
 Oxford is one, its shield being supported by a beaver 
 and an elephant) are among the exceptions. 
 
 The shield of arms, with the crested helmet above, the 
 motto beneath, and the supporters (if any) on either side, 
 form what is known as a hatchment or achievement of 
 arms. After the Reformation the hatchment of the 
 sovereign was ordered to be set up in every parish church 
 in token of the Act of Supremacy. In the church at 
 Stadhampton the shield of Queen Elizabeth still remains, 
 with the motto ' Reginae Nutrices Erint '. Usually when 
 the sovereign demised, his shield was removed and that 
 of his successor substituted. The custom continued 
 down to the reign of George III (probably the reputation 
 of his heir made the most loyal churchwardens hesitate 
 
266 THE GRAMMAR OF ARCHITECTURE 
 
 to set up his shield in their church) and his hatchment 
 still remains in many of our local churches. 
 
 Following the example of his sovereign the squire often 
 ordered that his shield should be set up in the village 
 church at his death. At Wood Eaton there are several 
 eighteenth-century hatchments, memorials to departed 
 squires, and at Nuneham and Besilsleigh the custom is 
 still retained. In Oxford, too, the home of forsaken beliefs, 
 it is still the practice when the head of a college dies to 
 set his shield above the gateway, there to remain until 
 his successor assumes office. 
 
 CHAPTER XIV 
 
 THE INTERIOR ARRANGEMENTS IN MEDIAEVAL 
 BUILDINGS 
 
 IN the Middle Ages there was only one type of building, 
 and it was made to serve with but slight modifications 
 as castle, house, and church, and even as barn. It is 
 the type represented to-day by the church nave and the 
 college hall. Its adjunct, which was also common to 
 buildings of whatever nature, was the fortified tower. 
 It was only in its interior arrangements * that this basic 
 type varied according to the purpose which the building 
 was designed to serve. I shall give a short account of 
 those arrangements in domestic, military, and ecclesiastical 
 buildings. 
 
 1 The ornamental details were, of course, more highly elaborated 
 in churches, but even they were of a single type. Early English 
 foliage may be seen in the shafts of the doorway of Appleton 
 Manor-house, and the ball-flower ornament in the doorway of 
 Fyfield Hall. 
 
MEDIAEVAL BUILDINGS 267 
 
 I. THE MANOR-HOUSE 
 
 The germ or nucleus of the house is the hall ; it is 
 significant that the word is still used to describe alike 
 the many-roomed mansion of the rich and the narrow 
 passage into which the front door of the cottage opens. 
 
 The earliest English house we do not refer to the 
 huts of peasants was a great hall like a church nave, 
 in the centre of the earthen floor was a stone hearth, the 
 smoke from which curled up among the rafters and found 
 its way out by a louvre, such as may still be seen in the 
 roof of Lincoln Hall. The cooking for the household 
 was done either in the open air or in a separate building 
 at one end of the hall ; at Stanton Harcourt the mediaeval 
 kitchen still remains ; it has no chimneys, and the smoke 
 from its fires escaped by means of shuttered openings in 
 the eaves, opened or closed according to the direction 
 of the wind. At the end of the hall, remote from the 
 kitchen, was a low platform or dais, upon which was set 
 the high table for the lord and his family. The lower 
 end of the hall was the province of the servants and 
 retainers, who took their meals on trestled tables. A 
 door near the dais opened upon an exterior flight of 
 steps leading to an upper chamber, the solar, built 
 against the gable-end of the hall and having below it 
 a storehouse or stable. To this secure and private 
 elevation the lord and his lady with their children 
 retreated at night, leaving their servants to sleep upon 
 the benches or the rushes of the hall. 
 
 Such was the ' house ' of the Norman baron ; and even 
 in the days of Edward I the king and queen gave audience 
 in the solar, seated upon their bed. 
 
 In the fourteenth century new needs produced develop- 
 
268 THE GRAMMAR OF ARCHITECTURE 
 
 ments in house-building ; the hall was not affected, but 
 the arrangement of buildings at its two ends became more 
 specialized. At the servants' end there grew up kitchens, 
 pantries, butteries, and larders, approached by doors in 
 the wall of the hall ; a wooden screen eight or ten feet 
 high protected the hall from draughts from these doors, 
 and a platform above it formed a gallery for minstrels. 
 This relationship between the hall and the kitchens and 
 buttery may still be seen in the older colleges, particularly 
 at New College. 
 
 At the lord's end of the hall the solar became sub- 
 divided into parlour and bedchamber. Then, as the 
 desire for privacy grew, more rooms were built on at 
 right angles to the hall ; and a similar development taking 
 place at the other end, three sides of a square were formed. 
 It only remained to build a wall with a gatehouse on the 
 remaining side and the quadrangular plan of the fourteenth- 
 century manor-house came into existence. This remained 
 general until Tudor times, when cannon made fortified 
 houses futile, and livery laws, by limiting the numbers 
 of personal retainers, made the great hall unnecessary. 
 
 In the Elizabethan house, therefore, the hall, though 
 still remaining the central core of the building, was much 
 reduced in size ; it was usually approached by a pro- 
 jecting porch in the middle of the front of the house, and 
 this, with the two wings representing the sides of the 
 mediaeval quadrangle, gave an E-shape to the plan which 
 is fancifully supposed to be meant as a compliment to 
 the queen. 
 
 Though fireplaces had been common enough in the 
 solars of mediaeval houses, it remained for the Tudor 
 builders to invent the chimney-stack ; in the earlier fire- 
 places the smoke escaped by means of flues in the thickness 
 
FIG. 115. THIRTEENTH-CENTURY FIREPLACE, ABINGDON ABBEY 
 
2/o THE GRAMMAR OF ARCHITECTURE 
 
 of the walls, and its vent at their outer face was hidden 
 in the angle of a buttress or turret. There is a fine 
 thirteenth-century fireplace in the ruins of Abingdon 
 Abbey, which has a lofty turret above it with concealed 
 openings for the escape of the smoke. But when domestic 
 architecture became definitely differentiated from ecclesi- 
 astical the necessity for chimneys was openly acknowledged 
 in building ; they were added to the fireplaces in older 
 houses ; it was at this time that the tall chimneys of 
 All Souls were built. 
 
 One proof that Late Gothic was not a debased style 
 is the artistic treatment of the new feature. People 
 who live in houses topped by groups of drain-pipes 
 should not cast reflections upon the builders of the 
 Jacobean chimney-stacks. 
 
 II. THE TOWER 
 
 The early castle differed from the manor-house only 
 in the greater importance of its strategic position and 
 in the greater strength of its defences. A castle was simply 
 a house, exceptionally well fortified. Down to the end 
 of the eleventh century its principal defence was the 
 moated mound with its wooden stockade surrounding 
 a circular area at the summit, within which the great 
 hall of timber was built. Early in the twelfth century 
 the castle hall was constructed of stone, and defended, not 
 by a stockaded mound, but by a great stone tower or 
 ' keep ' built near it. In the fourteenth century similar 
 towers were built in the wall that encircled the castle 
 yard. At Oxford these were destroyed when the castle 
 was ' slighted ' after the Civil War, but the Norman keep 
 still remains and is typical of the mediaeval tower. 
 
MEDIAEVAL BUILDINGS 
 
 271 
 
 Its immense walls of rubble are so thick that no batter- 
 ing of mediaeval artillery could seriously damage them ; 
 its only door was twenty feet from the ground, and when 
 the wooden stair by which it was reached was removed 
 it was impregnable, 
 and its garrison could 
 only be reduced by 
 starvation. The base- 
 ment space below the 
 first floor was reached 
 by a trap-door and 
 served as a store- 
 house. The first floor 
 was used as a hall, 
 and the floor above 
 as a chamber for the 
 women. In later 
 keeps small private 
 chambers were made 
 by hollowing the 
 thick walls ; fire- 
 places were similarly 
 made, the smoke 
 escaping by flues in 
 the wall. In times 
 of peace the keep 
 served as a private 
 dwelling for the lord 
 and his family, to 
 which they might retire from the crowd in the great hall. 
 It safeguarded the Norman family in the midst of an alien 
 and conquered race, and enabled a handful of Normans to 
 dominate the town or district that formed their holding. 
 
 FIG. 116. CHIMNEY TURRET, ABINGDON 
 ABBEY (c. 1250) 
 
272 THE GRAMMAR OF ARCHITECTURE 
 
 The top story of the tower and its flat roof were the 
 scene of defensive operations when the castle was attacked. 
 The arched openings in its four faces were closed by 
 wooden doors, hinged at the bottom, which dropped 
 outwards to form platforms, so that men standing upon 
 them could drop stones, boiling water, or molten lead upon 
 the heads of those who tried to plant scaling-ladders 
 against the wall or to attack its base with pickaxes. Later 
 in the twelfth century the parapets of towers were 
 pierced with holes for projecting beams to support 
 a wooden gallery for the same purpose. But this was 
 sometimes set on fire by the besiegers, and in the thirteenth 
 century overhanging galleries, carried on corbels, were 
 built of stone, having holes called machicolations pierced 
 in their floors. 
 
 In the north face of S. Michael's tower, i.e. on the side 
 most exposed to attack, is an arched opening like those 
 in the top story of the castle ; it is thirty feet from the 
 ground, and its purpose, like theirs, was to give access 
 to a projecting platform. The tower of S. Peter's as it 
 exists to-day shows in most of its details the work of the 
 fifteenth century. But since the original church was 
 built by Robert d'Oilgi and was always regarded as 
 guarding the city on the east : 
 
 Invigilat 'portae Australi Boreaeque Michael, 
 Exortum solem Petrus regit atque cadentem, 
 
 there must have been a Norman tower, and the very 
 marked ' batter ' or inward slope of its walls, in which 
 it resembles the castle, suggests that the present tower, 
 in spite of subsequent alterations, is substantially a part 
 of the original fortifications of Oxford though it does 
 not stand upon the actual line of the wall. New College 
 
MEDIAEVAL BUILDINGS 273 
 
 tower with its loop-holed and windowless lower stories is 
 obviously designed mainly for the defence of the wall 
 from which it projects. But the other towers of Oxford 
 have little that is distinctively military about them, 
 though they could always be used as temporary strong- 
 holds, impregnable to anything short of siege artillery. 
 
 III. THE CHURCH 
 
 The church stands upon land either given by some 
 Saxon thane or Norman lord, or else once forming part 
 of the open space round which the houses of the Saxon 
 hamlet clustered, and on which the folk-moots and the 
 markets were periodically held. In the former case church 
 and manor-house are usually near neighbours, e. g. Holy- 
 well ; in the latter the church stands in the central area 
 from which the streets radiate, e. g. S. Martin's, Carfax. 
 
 The churchyard was entered by a lych-gate, sheltered 
 by a timber roof, beneath which the bearers rested for 
 a moment in carrying a villager to his last sleeping-place. 
 Few lych-gates remain there is one at Garsington but 
 the little modern gateway of S; Martin's at Carfax serves 
 to remind us of them. The church always stood on the 
 north side of the churchyard, as is well seen at S. Giles's 
 and S. Mary Magdalene ; primitive people were sus- 
 ceptible to the influence of light and shade ; they 
 shuddered at the idea of burial in the shadow of the 
 church. 
 
 We have seen that the principal entrance to the building 
 was by the south door. Near it (as at Headington) stood 
 the cross, raised upon steps, symbolical of Calvary ; some- 
 times it stands still upon the very spot where the mis- 
 sionaries of Augustine or Birinus planted their wooden 
 
 785 S 
 
274 THE GRAMMAR OF ARCHITECTURE 
 
 crosses and first preached to the sons of Woden the 
 Gospel of Christ, and Him crucified. In the Middle 
 Ages the parson seldom preached, and never unless he 
 had something to say ; then, as a rule, he stood on the 
 steps to deliver 'his message, for pulpits were very rare 
 until after the Reformation. The outdoor pulpit at 
 Magdalen reminds us of the ancient practice. 
 
 In the church porch parish business was transacted, 
 and parts of the baptismal and marriage services were 
 conducted ; it was provided with stone seats, and con- 
 tained a stoup for holy water in which the entering 
 worshippers dipped their fingers. The Norman stoup at 
 S. Peter's has been hacked away, but its traces remain, 
 and recesses for stoups may be seen at S. Giles's Church 
 and All Souls Chapel. The room above the porch, 
 sometimes called the parvise, was used by the priest or 
 sacristan, or possibly in some cases was the cell of an 
 anchorite. After the Reformation it was often utilized 
 as a parish library ; for every church was obliged to 
 possess certain specified books defining the doctrines 
 of the Protestant faith. 
 
 The font stood either in the porch or immediately 
 within the church door, for the unbaptized had no right 
 of entry. Its form and ornaments varied with the 
 architectural fashion of its date. 
 
 Early fonts were frequently square, and were usually 
 supported on pillars ; the Norman font at IfHey and the 
 Early English font of S. Giles's are good examples. 
 S. Peter's font is a modern imitation of a Norman form 
 of which an original example may be seen at Radley. 
 Plain tub-shaped fonts, standing on great square plinths, 
 were common in parish churches in the twelfth and 
 thirteenth centuries ; there are specimens at Cowley and 
 
MEDIAEVAL BUILDINGS 275 
 
 Elsfield. After the thirteenth century the chalice form 
 became the most usual ; the cup was polygonal, and its 
 six or eight faces were panelled with blank tracery, or 
 carved with shields of arms. The font of S. Mary 
 Magdalene belongs to the middle of the fourteenth 
 century, that at All Saints', which came from S. Martin's 
 Church, is a little later. Fifteenth-century fonts are 
 smaller than those of earlier dates, for the original 
 practice had been totally to immerse the unfortu- 
 nate infant ; the font of S. Michael's will serve as an 
 example. 
 
 Inside the church the most striking object was the 
 great carved screen that separated the nave from the 
 chancel. The nave had many secular uses in the Middle 
 Ages ; it seems to have served most of the purposes of 
 a parish room ; its floor area was free from seats, except 
 for stone benches round the walls which may still be 
 seen in Merton Chapel and Cuddesdon Church ; and 
 in it were kept various articles of public property ; the 
 firehook, with which the thatched roof of a burning 
 cottage was pulled bodily off to prevent the spread of the 
 fire, the public coffin, in which the bodies of the villagers 
 were carried to their graves for wood was not wasted 
 on coffins for the poor, the arms which every village was 
 bound to provide according to the number of its able- 
 bodied men, the whip of the functionary who drove 
 out the dogs, the long wand of the sluggard-waker, the 
 ducking-stool for militant females, and even the copper 
 pot in which beer was brewed for the ' church ale ', a 
 mediaeval method of raising funds by the sale of ale 
 brewed by church workers. The screen, therefore, had 
 an important significance. 
 
 It was usually made of oak, but occasionally of stone, 
 s 2 
 
276 THE GRAMMAR OF ARCHITECTURE 
 
 as at Balking ; its top formed a platform, the Rood- 
 loft, from which parts of the mass service were said 
 or sung ; and it supported the great Crucifix or Rood 
 from which screen and loft took their names. All 
 the roods and most of the rood-lofts perished at the 
 Reformation, but many screens still survive, though 
 mutilated. 1 In the church of the Cowley Fathers all has 
 been revived, even to the replacing of the organ in the 
 rood-loft, and the practice of singing anthems from its 
 platform. The worn condition of the steps by which 
 the rood-loft was reached, and which still exist in most 
 of the village churches, at IfHey for example, is evidence 
 of the constant use of the rood-loft in the Middle Ages. 
 
 The walls were covered with painted illustrations of 
 Christian truths ; over the chancel arch was a representa- 
 tion of the Last Judgement : the dead arising naked 
 from their graves ; on one side, the gates of hell wide 
 opened for the wicked, with Satan and his demons 
 dragging them in; on the other, the just, looking upward ; 
 and above, Christ throned in glory. Opposite the church 
 door was a fresco showing S. Christopher bearing the 
 Infant Jesus ; before this the intending traveller prayed 
 to be safeguarded on his journey, that he might come 
 home again in peace. In the splay of the windows 
 were painted the pictures of the saints, and the glass itself 
 set forth their story. 
 
 Even good Protestants may view with regret the coats 
 of whitewash that cover these pictures, and the cold print 
 of the creed and ten commandments that replaced them 
 after the Reformation. In Oxford itself funds have been 
 
 1 There are fine fifteenth-century examples at Church Han- 
 borough and Charlton, and one of the thirteenth-century at 
 Stanton Harcourt. 
 
MEDIAEVAL BUILDINGS 
 
 277 
 
 found for the modern restorer to scrape whitewash, 
 plaster, and pictures together from the walls ; but in the 
 
 FIG. 117. WALL-PAINTING, SOUTH LEIGH 
 S. Michael weighing a Soul. 
 
 village churches traces of mediaeval colouring may still 
 be seen ; the restored wall-paintings at South Leigh are 
 almost as famous as the stained glass of Fairford. There 
 
278 THE GRAMMAR OF ARCHITECTURE 
 
 are faint traces of frescoes on the ceiling of the Chapter 
 House. 
 
 The pulpit may almost be said to be a post-Reformation 
 institution, though there are mediaeval examples at 
 Hanborough (in wood) and Combe (in stone) ; these, 
 however, belong to the fifteenth century, and are 
 probably due to the influence of Wicliffism. Pulpits 
 may be said to date from the reign of Edward VI, when 
 all churches were ordered to be so provided. James I 
 re-issued the injunction evidently with more effect, for 
 pulpits of his date are still very numerous. That in the 
 Cathedral is a fine example. In many churches the nine- 
 teenth-century craze for Gothic resulted in the destruc- 
 tion of the Reformation pulpits, and their replacement by 
 stone ones with Gothic details such as may be seen in 
 S. Mary's. The high pews with which church naves 
 were furnished after the Reformation were also destroyed 
 by the Victorian restorers, and replaced by the modern 
 seats on the model of the benches introduced into a few 
 churches in the fifteenth century. 
 
 In the chancel the principal feature was, of course, the 
 high altar. This was always of stone in the Middle Ages, 
 and was marked with five crosses symbolical of the Five 
 Wounds. Behind it was the carved reredos, with its 
 canopied recesses and figures, of which New College and 
 Magdalen possess unequalled examples. At S. Michael's 
 the fifteenth-century reredos of the high altar has been 
 moved to the chapel ; that at S. Mary's still remains, 
 but has been much mutilated. Where there was no 
 carved reredos there was often, as in the south aisle of 
 Holywell (Fig. 56), niches or brackets for the figures 
 of the Blessed Virgin and S. John, witnesses of the 
 Sacrifice commemorated at the altar. In order that 
 
MEDIAEVAL BUILDINGS 
 
 279 
 
 worshippers in the aisle might view the elevation of the 
 Host at the high altar a hole, called a squint, was cut 
 through the wall at the side of the chancel arch. At 
 Haseley there are squints giving sight of the altar from 
 both aisles of the church. 
 
 FIG. 118. SQUINTS AT HASELEY 
 
 An ever present accessory to the altar was the small 
 drain, called a piscina, in the south wall, in which the 
 chalice was washed after the celebration of mass. Its 
 arched recess, moulded according to its date, usually 
 contains also a stone shelf or credence upon which the 
 sacred vessels stood. These, when not in use, were kept 
 
280 THE GRAMMAR OF ARCHITECTURE 
 
 in a locker, a stone cupboard in the thickness of the wall ; 
 its wooden door has usually disappeared, but the iron 
 staples for hinges often remain, and the hole into which 
 the bolt was shot. The presence of piscinae in other 
 parts of the church, e.g. at the east ends of the aisles, 
 as in S. Giles's, proves that altars once stood near ; all 
 through the Middle Ages it was the custom to leave 
 money for the endowment of chantries, i.e. altars at 
 which masses might be said for the souls of the testators. 
 The priest so supported was usually not he of the parish, 
 but one who lived by celebrating masses in chantries, 
 a ' mass priest '. Some chantries were supported by guilds 
 for the good of the souls of departed members. There are 
 traces of half a dozen chantries in Kidlington Church. 
 
 South of the high altar are three canopied seats 
 sedilia, in which sat the celebrant and his assistants. Those 
 at Merton are among the most beautiful examples in 
 England. In the opposite wall was often another canopied 
 recess, the Easter Sepulchre, in which the Crucifix of the 
 high altar was hidden on the evening of Good Friday, 
 to be brought out again with high ceremonial on the 
 morning of Easter Day. There is a fine example at 
 Stanton S. John. The sanctuary is usually divided from 
 the western part of -the chancel by a low carved railing. 
 This is generally of seventeenth-century work, for Arch- 
 bishop Laud issued an order that the altar should be 
 fenced off with rails ' so close together that dogs might 
 not pass between'; these are copied from Classic models ; 
 often they are twisted like the columns of S. Mary's 
 porch. Cumnor has fine examples of rails, pews, and 
 pulpit of this date. 
 
 In the chancels of many village churches, e. g. Cowley 
 (Fig. 97), there is one feature which has never yet been 
 
ill 
 
 
 Vt 
 
282 THE GRAMMAR OF ARCHITECTURE 
 
 satisfactorily explained ; it is a small low window in the 
 south-west corner, which shows by the hinges in its jambs 
 that it was closed by a shutter and not glazed. It was once 
 believed that such windows were used in communicating 
 lepers ; but lepers were not permitted even to approach 
 the villages, and moreover, these low-side windows are so 
 common that some more general need must have brought 
 them into existence. What that need was no one knows. 
 Mr. Christopher Markham in his monograph on the sub- 
 ject has examined fourteen theories, and has given reasons 
 for discrediting them all. The subject still remains the 
 ' Sphinx of Archaeology ' ; but the blocked low-side 
 window at Elsfield with its book-rest cut in the jamb 
 seems to point unmistakably to its use in some service 
 in which the officiating priest was inside the church, 
 and the worshipper or worshippers without ; a low- 
 side window at Wigginton, which has a canopied seat 
 in its jamb, points to the conclusion. 
 
 On the north-western pier of S. Giles's tower may still 
 be seen a large cross painted on the face of the stone. 
 It marks one of the points at which the building was 
 anointed with oil by Bishop Hugh of Lincoln when he 
 consecrated the church in the time of Cceur de Lion. 
 There were twenty-four of these crosses on the stone- 
 work of every mediaeval church, twelve within and 
 twelve without ; sometimes they were carved, as on 
 the southern jamb of the west doorway at Iffley, but they 
 must usually have been painted or they would scarcely 
 have perished so completely. 
 
 The tombs in our ancient churches have a great wealth 
 of interest, artistic, historical, and heraldic, but I can 
 here only give a very short account of the marks by which 
 the date of a monument may be judged. The earliest 
 
MEDIAEVAL BUILDINGS 283 
 
 tombs still remaining in local churches belong to the 
 thirteenth century. At this time the stone coffin was 
 generally sunk into the floor of the church and covered 
 either with a slab of stone or Purbeck marble, on which 
 was cut a large floriated cross enriched with Early Gothic 
 foliage, or with a cross-legged figure of the deceased, in 
 the chain-mail of the period, with kite-shaped shield, 
 and with sword half-drawn from the sheath. There are 
 examples of floriated grave-slabs in the Cathedral, and 
 of mailed knights at Dorchester and Haseley. 
 
 FIG. 120. FLAT GRAVESTONE, GREAT MILTON, OXON. (c. 1250) 
 
 Before the end of the thirteenth century the altar- 
 tomb became the most common type ; a stone table was 
 raised above the grave and upon its flat top, which actually 
 formed an altar at which masses were celebrated for the 
 salvation of the departed, rested his effigies ; the figures 
 were sometimes carved in stone as in the tomb of Sir 
 George Nowers (1425) in the Cathedral, but more often 
 were cut out of plates of a metal called laten, a mixture 
 of copper and tin ; they were then let into the face of the 
 stone and fastened with pitch and rivets ; Adam de 
 Brome's tomb in S. Mary's, from which the brass has 
 been torn, will illustrate the method. The finest brasses 
 in Oxford are in the floor of the ante-chapel at New 
 
284 THE GRAMMAR OF ARCHITECTURE 
 
 College, and commemorate some of the early Wardens. 
 The earliest brass is that of Warden Bloxham, 1387, in 
 Merton College Chapel. 
 
 FIG. 121. TOMB, WATERPERRY, OXON. (c. 1400) 
 
 Brasses and altar-tombs remained in fashion until the 
 middle of the seventeenth century, but after the Reforma- 
 tion, when pews were everywhere introduced into 
 churches, the mural monument became the most common 
 
MEDIAEVAL BUILDINGS 285 
 
 type. A half-length figure of the deceased was set up 
 in a recess in the wall above his marble gravestone 
 in the floor. Shakespeare's monument at Stratford is 
 an example known to every one. There are similar 
 monuments in S. Mary Magdalene and S. Michael's 
 Churches, in the transepts of Merton (Figs. 57 and 69), 
 and in the Cathedral. 
 
 In the eighteenth century most monuments were fixed 
 to the walls, and generally consisted of a large white 
 marble slab on which was engraved a long Latin eulogy 
 of the dead man," sometimes surmounted by an urn or 
 a bust. The monuments of the Spencers at Yarnton are 
 the finest examples in the district. 
 
 The poor were buried in the churchyard, and their 
 graves are seldom marked by any permanent memorial. 
 But in the sixteenth century, as the wealth of the nation 
 increased, the rise of a lower middle class is seen in the 
 churchyard stones bearing dates of Elizabeth and James. 
 These are much thicker than modern tombstones ; their 
 tops are cut into ogee-curves, and they are sometimes 
 ' decorated ' with skulls and cupids, or with festoons of 
 fruit and flowers in the debased Classic taste of their date. 
 There are good examples in the churchyard of S. Peter's, 
 and others at Iffley, Cowley, Headington, and indeed in 
 every ancient churchyard. 
 
 The inscriptions on tombstones open a wide field of 
 interest to the antiquarian but I have already wandered 
 far from the province marked out in the preface, and 
 have strayed from architecture into archaeology. I will 
 quote but one sentence, from an epitaph in North Hinksey 
 Church, which may not be out of place at the end of 
 this book : * Reader, look to thy feet : honest and loyal 
 men are sleeping under them.' 
 
INDEX 
 
 Abacus 24, 30, 152 
 
 Ashmolean 21, 22 
 
 Abingdon 193, 270 
 
 Ashmolean, Old, 231 
 
 Abutment 34, 131 
 
 P* r\ 
 
 Acanthus 30, 151 
 
 .Badges 203 
 
 Adderbury 215 ^T" /u ' AU *' ^' i 
 
 Akeman Street 20 
 Aldrich 78 
 Alfred 261 
 All Saints' 78, 99, 231, 
 
 Ball-flower 247 
 Baluster 198 
 Bampton 247 
 Base 157 
 
 All Souls 40, 79, 99, 
 122, 171, 195, 269, 
 
 Basement-course 183 
 Battlements 168 
 Beak-head 242 
 
 274 
 Altar 278 
 Altar-rails 280 
 Appleton 151, 266 
 Apse 43, 76, 94 
 Architects : Barry, 
 
 Beckley 97, 148 
 Besilsleigh 266 
 Bicester 152, 199 
 Billets 179, 244 
 Blenheim Palace 78, 
 
 Q ~ 
 
 Sir C., 86 ; Blom- 
 field, Sir A., 88; 
 Bodley, Mr., 88, 
 214 ; Buckler, Mr., 
 86 ; Butterfield, 
 Mr., 86 ; Champ- 
 
 OO 
 
 Bloxham 215 
 Boss 118 
 Botanic Gardens 74 
 Bowtell 245 
 Brasenose 76, 171, 
 
 neys, Mr., 88 ; 
 Cockerell, Prof., 30, 
 84 ; Hare, Mr., 88 ; 
 Hawksmoor, 78 ; 
 Jackson, Mr., 88; 
 
 Brasses 283 
 Brick 175 
 Broughton 215 
 Butler 261 
 
 Jones, Inigo, 74, 
 
 Camera 41, 78, 184 
 
 231 ; Palladio 74, 
 
 Capital 148 
 
 76, 231 ; Rickman 
 
 Carrel 102 
 
 56 ; Vanbrugh, Sir 
 John, 78; Water- 
 
 Cassington 201 
 Castle 49, 52, 270 
 
 house, Mr., 86; Wren 
 
 Cathedral 39, 43, 52, 
 
 76, 131, 135, 231 ; 
 
 58, 62, 65, 98, 100, 
 
 Wykeham 99, 103, 
 
 113, 116, 128, 142, 
 
 223, 244 145, 151, 154, 164, 
 
 Architrave 22 171, 193, 204, 208, 
 
 Arms, 257 
 
 215, 219, 234, 239, 
 
 Arris 24 
 
 246, 247 
 
 Ceiling 171 
 
 ,278 
 
 Chevron 242, 258 
 Childrey 236 
 Chimney 269 
 Chinnor 213 
 Christ Church : Can- 
 terbury Gate 80 ; 
 
 Dome 78; Hall 68; 
 
 Library 100, 231 ; 
 
 Meadow Buildings 
 
 86 ; Stairway 68, 
 
 122 
 Clarendon Building 
 
 22, 26, 78, 138 
 Clear-story 106 
 Clifton 148, 158 
 Cloister 100 
 Combe 278 
 Composite Order 32 
 Corbel 164, 182 
 Corinthian Order 30 
 Cornice 22 
 Corpus Christi 170 
 Cowley 129, 138, 150, 
 
 188, 228, 274, 280 
 Coxwell 1 68 
 Credence 279 
 Crest 262 
 Crocket 252 
 Crypt 43, 98 
 Cuddesdon 133, 151, 
 
 243, 247 
 Cumnor 74, 138, 186, 
 
 234, 247, 260, 263, 
 
 280 
 
 Curtain- wall 166 
 Cusp 212 
 
INDEX 
 
 287 
 
 Dais 267 
 
 Hatchment 265 
 
 Mansard 170 
 
 Dentils 28 
 
 Headington 186, 273 
 
 Marston 228 
 
 Divinity School 66, 
 
 Hertford 32, 88, 232 
 
 Merton 32, 62, 76, 88, 
 
 99, 125, 130, 135, 
 
 Hinksey 138, 228, 
 
 102, 120, 129, 155, 
 
 174, 193, 224 
 
 285 
 
 158, 160, 165, 177, 
 
 Dog-tooth 60, 243 
 
 Holton 151 
 
 180, 182, 193, 213, 
 
 Doilgi 49, 54, 258, 
 
 Holy Trinity 88, 232 
 
 217, 219, 228, 254, 
 
 264 
 
 Hood-mould 177 
 
 275, 284 
 
 Dome 41, 160 
 
 
 Metopes 26 
 
 Dorchester 171, 193, 
 
 Iffley 108, 126, 129, 
 
 Milton 109, 193, 215, 
 
 214, 223, 237, 247 
 
 142, 145, 150, 157, 
 
 283 
 
 Doric Order 25 
 
 161, 164, 177, 192, 
 
 Montagu 260 
 
 Drip-stone 177 
 
 201, 224, 274, 276, 
 
 Mouldings 238 
 
 
 282 
 
 Mullions 224 
 
 Easter Sepulchre 280 
 
 Inns 104 
 
 Mutules 26 
 
 Eaves 164 
 
 Ionic Order 28 
 
 
 Eglesfield 261 
 
 Islip 148, 151 
 
 Nave 95, 275 
 
 Elsfield 95, 151, 204, 
 
 
 New College 65, 79, 
 
 275, 282 
 
 Jackson, Mr., 88 
 
 130, 148, 156, 185, 
 
 Entablature 21 
 
 Jesus College 86, 171, 
 
 222, 224, 268, 272, 
 
 Ewelme 104, 148, 
 
 228, 264 
 
 278, 283 
 
 175 
 
 Joseph's Stone 20 
 
 Noke 228 
 
 Exeter College 70, 84, 
 
 
 Northmoor 136 
 
 86, 125 
 
 Keble 86, 176 
 
 Nowers, Sir G., 279 
 
 
 Keel- moulding 179, 
 
 Nuneham 266 
 
 Fairford 277 
 
 246 
 
 
 Fillets 28, 248 
 
 Kettel Hall 136, 195 
 
 Orders, the, 24 ; Co- 
 
 Foliage : Corinthian 
 30; Stiff-stalked 153; 
 
 Kidlington 174, 280 
 
 rinthian 30 ; Doric 
 25 ; Ionic 28 ; Tus- 
 
 Naturalistic 154, 
 
 Label 177 
 
 can 28 ; Composite 
 
 234, 250 ; Square 
 
 Lancet 202 
 
 32 
 
 155 
 
 Laten 283 
 
 Oriel College 70, 170, 
 
 Font 274 
 
 Leigh, North, 20 
 
 183, 230, 255 
 
 Forest Hill 186 
 
 Leigh, South, 277 
 
 
 Forster 260 
 
 Lewknor 213 
 
 Panel 184 
 
 Four-leaved flower 
 
 Lincoln College 32, 
 
 Parapet 164 
 
 247 
 
 70, 230, 232, 267 
 
 Parvise 274 
 
 Frieze 22 
 
 Lintel 16 
 
 Pediment 161 
 
 
 Locker 280 
 
 Pellets 243 
 
 Gable 161, 170 
 
 Long and short work 
 
 Piddington 212 
 
 Gargoyle 165 
 
 126, 174 
 
 Pilaster 36 
 
 Glass 219 
 
 Louvre 267 
 
 Pinnacle 66, 123, 129, 
 
 Guttae 26 
 
 Low-side window 282 
 
 173 
 
 
 
 Piscina 279 
 
 Hanborough 276, 278 
 
 Machicolations 272 
 
 Pole 224 
 
 Harcourt 258 
 
 Magdalen 68, 79, 88, 
 
 Porch 274 
 
 Haseley 157, 213, 
 
 173, 185, 195, 225, 
 
 Portcullis 264 
 
 237, 247, 279 
 
 230, 264, 274/278 
 
 Pulpit" 274, 278 
 
288 
 
 INDEX 
 
 Queen's College 40, 
 76, 78, 161, 184, 
 195 
 
 Radcliffe 170, 258 
 Radley 274 
 Reredos 278 
 Rewley 252 
 Ribs 113 ; Wall-ribs 
 
 118; Ridge-ribs 118; 
 
 Tiercerons 120 ; 
 
 Liernes 120 
 Romsey 108 
 Rood-screen 275 
 Rubble 174 
 Ruskin 86 
 
 S. Aldate's 99, 217 
 S. Andrew's 89, 230 
 S. Barnabas 88, 94 
 S. Clement's 84, 232 
 S. Cross, Holy well 
 
 52,98,148,154,186, 
 
 241, 278 
 S. Ebbe's 52, 188, 
 
 S. Giles's 60, 98, 145, 
 154, 159, 180, 181, 
 193, 204, 205, 209, 
 
 212, 214, 223, 244, 
 273, 274, 280, 282 
 
 S. John's College 183, 
 195, 228 
 
 S. John the Evan- 
 gelist 123, 214, 276 
 
 S. Martin's 212, 273 
 
 S. Mary Magdalene 
 65, 99, 129, 148, 157, 
 158, 168, 193, 217, 
 250, 273, 275 
 
 S. Mary the Virgin 65, 
 68,132,157,172,213, 
 214, 250, 252, 278 
 
 S. Michael's 46, 49, 
 126, 148, 166, 174, 
 198, 214, 216, 239, 
 272, 278 
 
 S. Mildred's 103 
 
 S. Paul's 30, 84, 94 
 
 S. Peter in the East 
 52, 65, 95, 97, 98, 
 108, 113, 150, 154, 
 161, 16*,, 179, 186, 
 192, 201, 204, 215, 
 216, 223, 242, 244, 
 272, 274 
 
 S. Peter le Bailey 88, 
 218 
 
 S. Thomas's 138, 190, 
 240 
 
 Sandford 201 
 
 Schools, New, 88 
 
 Schools, Old, 32, 195, 
 255 
 
 Scroll-moulding 180 
 
 Sedilia 248, 252, 280 
 
 Sheldonian Theatre 
 76, 231 
 
 Silchester 95 
 
 Solar 267 
 
 Spandrel 252 
 
 Squint 279 
 
 Stadhampton 265 
 
 Stairs 170 
 
 Stanton Harcourt 
 267, 276 
 
 Stanton S. John 212, 
 214, 280 
 
 Stoup 274 
 
 Sunningwell 74 
 
 Supporters 265 
 
 Sutton Courtney 175 
 
 186, 202 
 Symbolism 233 
 
 Tidmarsh 95 
 Tiles 176 
 Tombs 282 
 Town Hall 88 
 Tracery 206 
 Transom 65, 220 
 Triglyphs 24 
 Trinity College 79, 88 
 Tudor Flower 252 
 Tuscan Order 28 
 
 University College 70, 
 86,88, 102, 170, 183, 
 228, 230, 255, 262 
 
 University Museum 
 86, 88, 233 
 
 University Press 32, 
 94, no 
 
 Villiers 260 
 Volute 28, 152 
 Voussoirs 140, 231 
 
 Wadham 32, 39, 68, 
 70, 74, 183, 228, 230, 
 232 
 
 Wall-painting 276 
 
 Wantage 239 
 
 Waterperry 134, 284 
 
 Widford 20 
 
 Wiggington 282 
 
 Witney 134 
 
 Wood Eaton 171, 
 181, 204, 228, 266 
 
 Woodstock 207 
 
 Yarnton 70, 285 
 
 Oxford : Horace Hart, Printer to the University 
 

 UNIVEKSITY OF CALIFORNIA LIBRARY 
 BERKELEY 
 
 THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 Books not returned on time are subject to a fine of 
 50c per volume after the third day overdue, increasing 
 to $1.00 per volume after the sixth day. Books not in 
 demand may be renewed if application is made before 
 expiration of loan period. 
 
 OCT 9 
 
 T| 3.6192 
 
 4 ?# 
 
 JUL 29 
 
 301986 
 
 50m-7,'16 
 
l ! o I ' 
 
 GENERAL LIBRARY - U.C. BERKELEY 
 
 274121 
 
 ERSITY OF CALIFORNIA LIBRARY