Patterns in Stonework: The Early Church in Britain and Ireland: An introduction to ecclesiastical geology 9781407306001, 9781407321615

This work falls into two parts. In the first, the author undertakes a summary of his ecclesiastical geological research

242 46 84MB

English Pages [224] Year 2009

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Front Cover
Title Page
Copyright
Frontispiece
Abstract
Table of Contents
LIST OF TABLES
LIST OF FIGURES
ACKNOWLEDGEMENTS
IMPORTANT NOTE AND AN APOLOGY
CHAPTER ONE PREAMBLE: WHAT LEVELS OF UNANIMITY EXISTED IN EARLY ECCLESIASTICAL ARCHITECTURE IN THE BRITISH ISLES?
CHAPTER TWO MASONRY DETAIL AND STONEWORK IN ENGLISH EARLY ECCLESIASTICAL ARCHITECTURE
CHAPTER THREE EVIDENCE OF ‘PATTERNED’ MASONRY STYLES IN SCOTLAND AND THE ISLE OF MAN
CHAPTER FOUR A GEOLOGICAL REVIEW OF MASONRY STYLES IN IRISH EARLY ECCLESIASTICAL ARCHITECTURE
CHAPTER FIVE A CRITIQUE OF THE GEOLOGICAL ASPECTS OF THE SUPPOSED MONASTIC INFLUENCES IMPOSED ON EARLY IRISH ECCLESIASTICAL MASONRY STYLES
CHAPTER SIX THE ANTAE PUZZLE: A GEOLOGICAL INTERPRETATION FOR THE CONSTRUCTION OF ANTAE
CHAPTER SEVEN RESERVATIONS AND CONCLUSIONS
GLOSSARY
REFERENCES
Recommend Papers

Patterns in Stonework: The Early Church in Britain and Ireland: An introduction to ecclesiastical geology
 9781407306001, 9781407321615

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

BAR 496 2009

Patterns in Stonework: The Early Church in Britain and Ireland An introduction to ecclesiastical geology

POTTER

John F. Potter

PATTERNS IN STONEWORK

BAR British Series 496 2009 B A R

Patterns in Stonework: The Early Church in Britain and Ireland An introduction to ecclesiastical geology

John F. Potter

BAR British Series 496 2009

ISBN 9781407306001 paperback ISBN 9781407321615 e-format DOI https://doi.org/10.30861/9781407306001 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Frontispiece A simplified geological map of Britain and Ireland after the British Geological Survey. IPR/118-7ACY British Geological Survey © NERC. All rights reserved.

ABSTRACT Unusually, this work was motivated by two very unlike objectives. Various workers in the fields of church architecture, archaeology and history had requested that, for ease of reference, the author should condense the varied results of his studies in ecclesiastical geology into a single volume. At the time, the author’s brief analyses on the earliest Irish churches and their incorporation of stones set in vertically orientated fashion was proving to be unacceptable to the recognised ecclesiastical experts in that country. To them, the early practices of masons in Ireland stood apart from anything that might have been occurring elsewhere in the British Isles. Any detailed analysis of the stonework in Irish churches required the explanatory background already applied in England. This work, therefore, falls into two parts. In the first, the work of thirty plus years are briefly summarized, in the second part this information is applied to a number of early churches in Ireland. Chapters 1 and 2 examine the characteristics of stone emplacement as they apply in particular to the AngloSaxon churches of England. They illustrate how the craftsmen of this period used stone in certain structural features of their ecclesiastical buildings in distinctive styles, and how these styles may be distinguished from the work of the Norman or ‘Romanesque’ period that followed. They also provide details of the simplified nomenclature that has been devised to describe the distinguishing bedding orientations that can exist for stones emplaced in different wall structures. In Chapter 3, some of these same styles of stone emplacement, more recently identified in various early ecclesiastical sites in Scotland and the Isle of Man, are discussed. Occurring at much the same time as the distinctive Anglo-Saxon work in England the styles are described as ‘Patterned’. The reasons for the subtle differences in styles between Scotland and England (and between regions) are considered and attributed to the specific controls of geology and available rock type. The following Chapters (4 to 6) examine a selection of early ecclesiastical sites in Ireland. Stone emplacement patterns in some thirty plus Irish ecclesiastical buildings are carefully reviewed, particularly with reference to their quoins, antae, and arch jambs. Where a high proportion of the stones in these structures are set with their bedding or lineation set vertically, they replicate the ‘Patterned’ style observed in buildings in England, and more especially, Scotland. Those portions of buildings perceived as reflecting these patterns are considered to be of a similar early date, and the particulars of those structures exhibiting them are detailed. This additional information enables parts of many early Irish ecclesiastical buildings to be more precisely dated. Furthermore, with so many Irish churches constructed of hard, difficult to distinguish and utilize, Palaeozoic rock lithologies, it permits different areas or periods of wall fabric to be more readily discriminated. This is exemplified in those churches which possess antae, where the workmanship provided dates of both during and after the ‘Patterned’ period. Although the purpose for the construction of antae may never be definitely known, Chapter 6 offers a new hypothesis based on the visible evidence revealed in the wall fabrics. This proposes that they were constructed primarily for the defence of the vulnerable corners of simple single-celled churches. Ecclesiastical geology is still in its formative years. This study in itself leaves many ecclesiastical sites in the British Isles unexamined. The final Chapter of this work, therefore, establishes a number of broad conclusions and briefly discusses a few of the many areas of work still requiring attention.

i

CONTENTS List of Tables ........................................................................................................................................................... v List of Figures .........................................................................................................................................................vi ACKNOWLEDGEMENTS ................................................................................................................................. xxv CHAPTER ONE. PREAMBLE: WHAT LEVELS OF UNANIMITY EXISTED IN EARLY ECCLESIASTICAL ARCHITECTURE IN THE BRITISH ISLES? 1.1 Humble beginnings ........................................................................................................................................... 1 1.2 Early answers. ................................................................................................................................................... 1 1.2.1 The ecclesiastical ammonite.................................................................................................................... 1 1.2.2 Gravel as a building stone ....................................................................................................................... 2 1.2.3 The use of ferruginously-cemented gravel and sand in ecclesiastical buildings ..................................... 3 1.3 Ecclesiastical Geology – a new study evolves .................................................................................................. 4 1.4 Geology and architectural style. ........................................................................................................................ 6 1.4.1 Geological influences and the Anglo-Saxon styles ................................................................................. 6 1.4.2 Stone bedding orientations ...................................................................................................................... 6 1.5 What then of this preamble’s title? ................................................................................................................... 7 CHAPTER TWO. MASONRY DETAIL AND STONEWORK IN ENGLISH EARLY ECCLESIASTICAL ARCHITECTURE 2.1 Historical introduction ...................................................................................................................................... 8 2.2 Selection of stone in English early ecclesiastical architecture .......................................................................... 8 2.2.1 Changes in stone choice and influences of transport over time............................................................... 8 2.2.2 The consecration of ecclesiastical sites ................................................................................................... 9 2.2.3 Identification of ecclesiastical stone types ............................................................................................ 11 2.2.4 Superficial deposits ............................................................................................................................... 12 2.2.5 Re-use of stone and other materials ...................................................................................................... 13 2.2.6 Other building stone variables............................................................................................................... 14 2.3 Patterns in early ecclesiastical architectural stonework .................................................................................. 18 2.3.1 Patterns in early quoin stones of the London Basin .............................................................................. 18 2.3.2 Patterned stone orientation in Anglo-Saxon churches........................................................................... 20 2.4 A new stone orientation nomenclature............................................................................................................ 23 2.4.1 New quoin stone nomenclature ............................................................................................................. 23 2.4.2 A new nomenclature for pilaster-strip stones........................................................................................ 27 2.4.3 A new nomenclature for arch jamb stones ............................................................................................ 28 2.5 Stone quarrying and extraction ....................................................................................................................... 29 2.5.1 The quarrying of stone .......................................................................................................................... 29 2.5.2 Working stone for the construction of specialist Anglo-Saxon architectural features ..........................31 2.6 Cut backs or plaster stops?.............................................................................................................................. 33 2.7 Recognition of an Anglo-Saxon ‘Patterned’ style ........................................................................................... 36 2.7.1 Introduction ........................................................................................................................................... 36 2.7.2 Patterns in wall fabric stonework .......................................................................................................... 36 2.7.3 The ‘Patterned’ style ............................................................................................................................. 41 2.8 The change in decorative style. ....................................................................................................................... 42 2.9 Changes in the Anglo-Saxon ‘Patterned’ style over time ............................................................................... 43 2.9.1 Introduction ........................................................................................................................................... 43 2.9.2 At what time did the Anglo-Saxon masons introduce stones with vertically orientated bedding into church wall structures? .................................................................................... 44 2.9.3 At what time did Anglo-Saxon masons introduce cut backs in stone?.................................................. 44 2.9.4 Can Anglo-Saxon wall patterns be dated?............................................................................................. 45 2.10 Exceptions to the rule .................................................................................................................................... 45 2.11 Influences of rock types and lithologies........................................................................................................ 45 2.11.1 Introduction ....................................................................................................................................... 45 2.11.2 Analytical summary. ......................................................................................................................... 47 2.11.3 Rock types in other Anglo-Saxon ecclesiastical features .................................................................. 50 2.12 What of the Anglo-Saxon Kingdoms? .......................................................................................................... 52 2.13 The Norman ‘Romanesque’ masonry styles ................................................................................................. 53 2.13.1 Introduction ....................................................................................................................................... 53 2.13.2 Examples of the use of stone............................................................................................................. 53 2.13.3 Stone orientation and cut backs in Norman churches........................................................................ 53 ii

CHAPTER THREE. EVIDENCE OF ‘PATTERNED’ MASONRY STYLES IN SCOTLAND AND THE ISLE OF MAN 3.1 A Scottish enigma ........................................................................................................................................... 56 3.2 Available rock types in Scotland ..................................................................................................................... 56 3.3 A review of early Scottish churches................................................................................................................ 56 3.3.1 Introduction ........................................................................................................................................... 56 3.3.2 Further Scottish difficulties ................................................................................................................... 57 3.3.3 Churches created in the ‘Patterned’ style .............................................................................................. 57 3.3.4 Abernethy Round Tower, Perth and Kinross (NO 190 165). ................................................................ 64 3.3.5 Brechin Cathedral Round Tower, Angus (NO 596 601). ...................................................................... 65 3.3.6 St Magnus, Egilsay (HY 466 304) ........................................................................................................ 67 3.4 ‘Patterned’ churches in the Isle of Man.. ........................................................................................................ 74 3.4.1 Introduction ........................................................................................................................................... 74 3.4.2 St Patrick’s Isle, Peel............................................................................................................................. 75 CHAPTER FOUR. A GEOLOGICAL REVIEW OF MASONRY STYLES IN IRISH EARLY ECCLESIASTICAL ARCHITECTURE 4.1 Introduction ..................................................................................................................................................... 78 4.2 Stone availability in Ireland ............................................................................................................................ 79 4.3 Aspects of ‘Patterned’ workmanship in Ireland .............................................................................................. 82 4.4 ‘Patterned’ workmanship in Ireland ................................................................................................................ 82 4.4.1 Introducing the varieties ........................................................................................................................ 82 4.4.2 A new nomenclature for stones in antae ...............................................................................................83 4.4.3 Scottish practices repeated in Ireland .................................................................................................... 83 4.5 Examples of ‘Patterned’ style workmanship in early Irish churches. ............................................................. 84 4.5.1 Introduction ........................................................................................................................................... 84 4.5.2 Ardmore Church, Waterford (X 188 774)............................................................................................. 85 4.5.3 Loorha, North Tipperary (M 920 046) .................................................................................................. 89 4.5.4 Monaincha Abbey, North Tipperary (S 170 884) ................................................................................. 89 4.5.5 Ardpatrick, Limerick (R 637 208)......................................................................................................... 93 4.5.6 Killulta, Limerick (R 434 534). ............................................................................................................. 95 4.5.7 Kilrush, Limerick (R 556 567).............................................................................................................. 95 4.5.8 Teampall Molua (from Friar’s Island), now at Killaloe, Clare (R 701 728). ........................................ 96 4.5.9 Noughaval (small church), Clare (R 208 967). ..................................................................................... 98 4.5.10 Oughtmama (Ucht Mama) (west church), Clare (M 304 078) ............................................................ 99 4.5.11 Temple Cronan (Teampall Chronain), Clare (M 289 000)................................................................ 101 4.5.12 St Cronan, Tuamgraney, Clare (R 637 830)...................................................................................... 103 4.5.13 Killeenemeer, Cork (R 775 070). ...................................................................................................... 104 4.5.14 Labbamolaga (small church), Cork (R 764 176)............................................................................... 105 4.5.15 Kilgarvan, Kerry (W 013 734) .......................................................................................................... 107 4.5.16 Ratass, Kerry (Q 853 141)................................................................................................................. 108 4.5.17 Rattoo, Kerry (Q 878 336) ................................................................................................................ 109 4.5.18 Agha, Carlow (S 730 654)................................................................................................................. 110 4.5.19 Glendalough Cathedral, Wicklow (T 123 968) ................................................................................. 113 4.5.20 St Mary, Glendalough, Wicklow (T 122 968)................................................................................... 118 4.5.21 Temple Ciarán, Clonmacnoise, Offaly (N 010 306) ......................................................................... 120 4.5.22 Temple Kelly, Clonmacnoise, Offaly (N 010 306) ........................................................................... 122 4.5.23 Dulane, Meath (N 741 788)............................................................................................................... 122 4.5.24 St Cianan, Duleek, Meath (O 045 684) ............................................................................................. 123 4.5.25 St Columba, Kells, Meath (N 740 758) ............................................................................................ 123 4.5.26 St Feichin, Fore, Westmeath (N 510 704)......................................................................................... 127 4.5.27 St Mel, Ardagh, Longford (N 204 686)............................................................................................. 127 4.5.28 Drumacoo, Galway (M 396 168) ...................................................................................................... 129 4.5.29 St Cavan, Inisheer, Galway (L 986 025) ........................................................................................... 131 4.5.30 Kilgobnet, Inisheer, Galway (L 975 027).......................................................................................... 133 4.5.31 Teampall Benan, Inishmore, Galway (L 884 071). ........................................................................... 133 4.5.32 Teampall Bhreacáin, Inishmore, Galway (L 811 121) ...................................................................... 135 4.5.33 Teampall Chiaráin, Inishmore, Galway (L 873 104)......................................................................... 139 4.5.34 Teampall MacDuagh, Inishmore, Galway (L 823 104)..................................................................... 139 4.5.35 Teampall na Neeve (Naomh), Inishmore, Galway (L 824 103). ....................................................... 141 4.5.36 Temple Soorney, Inishmore, Galway (L 866 105)............................................................................ 141 iii

4.6 4.7

4.8 4.9

4.5.37 Kilmacduagh Cathedral, Galway (M 405 000) ................................................................................. 141 4.5.38 Kiltiernan (Cill Tiarain), Galway (M 437 156). ................................................................................ 143 4.5.39 Banagher, Derry (C 676 066)............................................................................................................ 145 Brief analysis of Irish churches described which possess ‘Patterned’ characteristics. ..................................148 The ‘Patterned’ characteristics of early Irish ecclesiastical buildings. .......................................................... 149 4.7.1 Vertically orientated bedding in stonework. ....................................................................................... 149 4.7.2 Architectural features of the ‘Patterned period ................................................................................... 149 4.7.3 Other non-definitive factors of the Irish ‘Patterned’ period ................................................................ 150 4.7.4 The Irish ‘Patterned’ period: its duration. ........................................................................................... 150 Ecclesiastical buildings of the Gallarus type................................................................................................. 152 Conclusions and implications. ...................................................................................................................... 153

CHAPTER FIVE. A CRITIQUE OF THE GEOLOGICAL ASPECTS OF THE SUPPOSED MONASTIC INFLUENCES IMPOSED ON EARLY IRISH ECCLESIASTICAL MASONRY STYLES 5.1 Introduction ................................................................................................................................................... 154 5.2 A brief review of the Ó Carragáin approach ................................................................................................. 155 CHAPTER SIX. THE ANTAE PUZZLE: A GEOLOGICAL INTERPRETATION FOR THE CONSTRUCTION OF ANTAE 6.1 Introduction................................................................................................................................................... 162 6.2 The evidence ................................................................................................................................................. 162 6.2.1 Evidence for possible roof support...................................................................................................... 162 6.2.2 The structural evidence. ...................................................................................................................... 164 6.3 The physical protection of early Irish churches ............................................................................................ 168 6.3.1 What was the need? ............................................................................................................................. 168 6.3.2 Megalithic random stonework.. ........................................................................................................... 169 6.3.3 Protecting a small church building – the use of antae.........................................................................170 6.3.4 The distribution of churches with antae..............................................................................................170 6.4 Other aspects of antae construction ..............................................................................................................171 6.4.1 Apparent frailty related to lack of wall bonding. ................................................................................ 171 6.4.2 Depth and size of antae.......................................................................................................................172 CHAPTER SEVEN. RESERVATIONS AND CONCLUSIONS 7.1 Reservations and outstanding problems.. ...................................................................................................... 175 7.1.1 Introduction ......................................................................................................................................... 175 7.1.2 Analysis of mortars ............................................................................................................................. 175 7.1.3 Were the external surfaces of ‘Patterned’ ecclesiastical buildings limewashed or plastered?. ....................................................................................................................................... 175 7.1.4 Areas of potential analysis – plinths.................................................................................................... 175 7.1.5 Changes in wall thickness. .................................................................................................................. 176 7.1.6 What of Irish Round Towers?. ............................................................................................................ 176 7.1.7 The west doorway – ‘A priest in residence?’ ..................................................................................... 179 7.1.8 Earlier and later stonework styles ....................................................................................................... 180 7.1.9 Architectural detail. ............................................................................................................................. 180 7.2 Conclusions................................................................................................................................................... 180 7.2.1 Introduction ......................................................................................................................................... 180 7.2.2 Conservation and preservation of early ecclesiastical buildings. ........................................................ 182 7.2.3 Great Britain, Ireland and beyond ....................................................................................................... 182 GLOSSARY ........................................................................................................................................................ 183 REFERENCES .................................................................................................................................................. 186

iv

LIST OF TABLES Table 2.1

An assessment of some stonework features which are believed to distinguish churches of ‘Patterned’ style as observed in England. These characteristics are in all instances dependent on the availability of rocks of suitable lithologies. .......................................................... 48

Table 3.1

‘Patterned’ stonework features recognised in three early churches in Scotland.................................62

Table 4.1

A summary of the principal Devonian and Carboniferous rocks of Ireland. It should be noted that some of the deposits shown are only regional in occurrence. Information partly from the Geological Survey of Ireland.................................................................................... 81

Table 4.2

An alphabetical list of the ‘Patterned’ churches in Ireland examined in this work ...........................84

Table 4.3

A simplified assessment of some of the stonework features which are believed to distinguish churches of the ‘Patterned’ (in England, Anglo-Saxon) style. This assessment applies to early churches in Britain and Ireland and in all instances is influenced by the availability of suitable rock lithologies. The information given here is based upon, and supplements, the details provided in Potter (2008c; 2008d)................................. 151

Table 6.1

A table illustrating the variable lengths which antae project from their respective gable walls. The breadth of each anta is also given together with the likely age of their construction as determined from their stonework ........................................................................... 173

Table 7.1

West doorway widths for various early Irish churches. Measurements are recorded to the nearest five millimetres. ........................................................................................................... 179

Glossary

A chrono-stratigraphic column for the rocks of Britain cited in this work. Details of the rocks referred to in Ireland are given in Table 4.1 and the text....................................................... 185

v

LIST OF FIGURES Frontispiece A simplified geological map of Britain and Ireland after the British Geological Survey. Published here by kind permission of the Natural Environment Research Council. Figure 1.1 The north side of the chancel of the church of St Mary, Ripley (TQ 052 566), in Surrey. The ferruginously-cemented blocks of gravel included in this rebuilt wall initiated the present author’s studies in ecclesiastical geology. The gravel blocks are mixed with nodular flints and the steel rule is approximately 320mm. in length ............................................................................................................................................................ 1 Figure 1.2 A sketch of an ammonite similar to the fragmentary form discovered in the walls of the chancel of Ripley church (x 1.3). ..................................................................................................................................................... 2 Figure 1.3 The route of an ecclesiastical ammonite: depicting the travels of an ammonite from the place of its first preservation as a fossil to its final resting place in the chancel of Ripley church in Surrey. ................................... 2 Figure 1.4 Ferruginous gravel in the wall of a gravel pit, now filled, near Thorpe, Surrey. ......................................... 3 Figure 1.5 An example of naturally concreted gravel from Linch Hill (SP 403 040), in Oxfordshire. Bridgland (1994, 7) correlates these gravels with those of the Taplow Terrace in the Middle Thames region. (Photo. courtesy, British Geological Survey, A12429). ............................................................................................................. 3 Figure 1.6 About 2km. north-west of Ripley church, Pyrford church (TQ 040 583) stands on a marked terrace above the River Wey. The major component of the church wall fabric is ferruginously-cemented gravel. In this instance, observed by a party in the west nave wall. The wall is supported by a later sarsen buttress. .......................... 4 Figure 1.7 The extensively rendered Wisley church (TQ 057 596) is about 3km. due north of Ripley church. In the south wall of the nave removed render reveals ferruginously-cemented gravel adjoining the structure of a later brick window. 45 Figure 1.8 In the London Basin the present author has examined the external stonework of all churches. The region of the Basin may be defined as the boundary between the Upper Cretaceous (Chalk) and the Palaeogene rocks; this is marked by a solid line. The broken line bounds the area in which all churches have been visited. Major national grid lines are marked. ............................................................................................................................ 5 Figure 1.9 Churches in the London Basin incorporating ferruginously-cemented gravel or sand in their fabric. The lack of frequency in the immediate region of London is believed to reflect the repeated rebuilding of churches in the region related to population growth since Anglo-Saxon times. The broken line indicates the area in which all churches have been examined. (Largely after Potter, 2001a)...................................................................... 5 Figure 1.10 The existing north transept to the church of St Peter and St Paul, Appledore (TQ 957 293). The north-east quoin, which in Anglo-Saxon times was probably the north-east quoin of the nave, is composed of ferruginously-cemented sandy-gravel blocks. Many of these blocks are set with their bedding orientated vertically, which the author is demonstrating to a party of people. ................................................................................ 7 Figure 1.11 Detail of the quoin illustrated in Figure 1.10, in which some of the stones exhibit bedding orientated vertically (pen provides scale). ...................................................................................................................................... 7 Figure 2.1 The tower of St Peter, Barton-on-Humber (TA 035 219) was one of the first examples of ecclesiastical architecture to be attributed by Rickman (1817) to be of pre-Conquest origin. It is here viewed from the south-west. ....................................................................................................................................................... 8 Figure 2.2 Clapham Tower (TL 035 525) was also determined to be of Anglo-Saxon age by Rickman (1817). Again viewed from the south-west, as at Barton the tower is surmounted by a younger, in this instance Norman, belfry. ............................................................................................................................................................................ 9 Figure 2.3 The Anglo-Saxon use of a) Barnack Stone and b) Quarr Stone in quoins of ecclesiastical buildings in the southern half of England: the details of these churches are presented in Potter (2005b). ........................................ 10 Figure 2.4 The south wall of the chancel at Greenford (TQ 145 832), in Middlesex, has principally been rebuilt in relatively modern times using cobbles of quarried and knapped flint. Ferruginously-cemented sand and sandygravel blocks from an earlier church, however, have been re-incorporated decoratively in the walls............................ 10 vi

Figure 2.5 Heckfield church (SU 723 605) in north Hampshire is here viewed from the north-west to show the north aisle. Very unfortunately, the south wall referred to in the accompanying text has now been covered in cement render, but the fabric in this north aisle is probably re-used similarly and obtained from the north nave wall. ................................................................................................................................................................................ 11 Figure 2.6 The tower of St Peter, Ash (SU 897 508) is constructed of sarsen stones: these stones from the detail of their working were used in building the tower about the 15th century (Potter, 1998, 295). ....................................... 12 Figure 2.7 Contoured distribution of those churches in the London Basin which exhibit silcretes in their fabric. The distribution has been weighted for quantity of silcretes, as follows: a significant amount (x3); some, more than 10 pieces (x2); less than 10 pieces (x1). Contouring has been based on 10km. diameter centres. ‘P’ indicates the concentration of Hertfordshire Puddingstone silcrete material and ‘S’ those concentrations of sarsens (irrespective of letter size). The map is based on statistics in Potter (1998). .................................................................. 13 Figure 2.8 A weighted and contoured density map of churches in the London Basin containing Roman tiles (from Potter, 2001b). A weighting of x2 was given for those churches with tiles used in a structural capacity (as quoins) and the distribution was based on 10km. centres. .............................................................................................. 14 Figure 2.9 Numerous stones from the rocks of the Charnwood Forest area of Leicestershire are included in the lower walls of Brixworth church (SP 747 713), in Northamptonshire. They are believed to have been incorporated from an earlier Roman structure, possibly one similar to the Jewry Wall, in Leicester (Sutherland, 2003, 114). The two dark rocks (left) are of Cambrian, Swithland Slate and the rock top right is an unusual mica rich hornfels, from the Charnian sequence. .................................................................................................................... 15 Figure 2.10 One of the many Roman stones re-used in Atcham church (SJ 541 092) walls (Potter, 2005f), this example displaying carving. ........................................................................................................................................... 15 Figure 2.11 Wroxeter church (SJ 563 082) possesses walls constructed from a variety of local rock types first used in the Roman site of Viroconium (Potter, 2005f). From the same source is the font which probably previously formed the base to a column (now inverted). ............................................................................................... 16 Figure 2.12 Burnt reddened stones from the Ashdown ‘Beds’ sandstone occur around the 13th century tower arch of St Mary, Kenardington (TQ 975 322). Blocks of Hythe Beds and dark ferruginously-cemented sandygravel are also present. In 1559, the nave of the church was struck by lightning and destroyed. (See also Figure 2.13 and Potter 2005d for a full description). ................................................................................................................. 16 Figure 2.13 Kenardington church viewed from the north-east places Figure 2.12 in context. Note that the nave was rebuilt to the south of the axial line of the earlier church. ....................................................................................... 16 Figure 2.14 The detail of a septarian nodule from the London Clay displayed in the south nave wall of St Mary the Virgin, Salcott (TL 952 137). The nodule (centre) has been bored by the bivalve Pholas, indicating its collection from a marine site such as a beach (20p coin provides scale). Below the nodule is a flint cobble. ............... 17 Figure 2.15 Detail of an encrusting marine serpulid on an Ashdown ‘Beds’ calcareous sandstone boulder on the north nave wall of the original Anglo-Saxon church at Lydd (TR 043 209), affording clear evidence of the boulder’s marine origin (Potter, 2005d). ........................................................................................................................ 17 Figure 2.16 The detail of boulders, collected from a marine environment, from the north-east arcade pier of the nave of St Augustine, Canterbury (TR 155 577). Immediately above the lens cap (50mm.) a block of Folkestone Beds sandstone has been encrusted by several serpulids. A grey slab of Hythe Beds calcareous sandstone has been bored by a Pholas, 50mm. to the right of the cap. ................................................................................................. 17 Figure 2.17 This London Clay septarian nodule (below pen) which occurs in the walls of Great Braxted church (TL 851 155) has Roman cement (opus signinum), with included brick fragments, adhering to its surface. A white ‘ring’ of calcium carbonate, representing the point of attachment of an oyster, partly covers both rock and Roman cement. The nodule must have been collected, therefore, from a Roman structure at a time when it was beneath the sea. The Roman shore fort (?Orthona), about 30 km. via the River Blackwater provides such a source. ............................................................................................................................................................................ 18 Figure 2.18 The south-west corner of the nave of St Botolph, Northfleet (TQ 624 741) displays stones of travertine set in long and short style with a number of the ‘long’ stones placed with their bedding vertically (Potter, 2006b). .............................................................................................................................................................. 19 vii

Figure 2.19 The 10th to the 12th travertine stones above ground level in the south-west quoin of the tower at Corringham (TQ 716 742) are illustrated. Stone twelve shows the vertical orientation of its bedding (with its face to the right) particularly well. The lens cap is 50mm. in diameter, and the church has been described in more detail by Potter (2005a). ............................................................................................................................................... 19 Figure 2.20 Much of the wall fabric of All Saints, Inworth (TL 879 178) is constructed of ferruginouslycemented gravel (Potter, 2005a). The church, here viewed from the north, was first noted for its Anglo-Saxon double-splayed windows by Chancellor (1904-5). Some of the original chancel quoin stones, built prior to the chancel extension, exhibit Anglo-Saxon, vertically orientated bedding. .................................................................... 19 Figure 2.21 Part of the north-west nave quoin of Birchanger church (TL 507 228). Each quoin block of Barnack Stone has been emplaced with its bedding vertical and the bedding planes facing to the right (BVFR). ...................... 20 Figure 2.22 The long and short style, south-west quoin to the nave at St Mary the Virgin, Strethall (TL 484 398), in Essex. ........................................................................................................................................................................ 20 Figure 2.23 Detail of the lowest three stones in the long and short style, south-west nave quoin at Strethall church of Figure 2.22. From the lowest stone upwards, the bedding orientations (see section 2.4.1) are: BVFR, BH, BVFL. .................................................................................................................................................................... 20 Figure 2.24 Part of the south-west quoin of the chapel at Bradwell-on-Sea (TM 031 082) constructed of megalithic stones re-used from the local Roman site. The Barnack Stone used displays the vertically orientated bedding well in the low sunlight. The second stone possesses the hole for a Roman cramp. ........................................ 21 Figure 2.25 The west face of Corbridge church (NY 988 644), Northumberland, here shows the south-west quoin of the original west porch to an Anglo-Saxon church, together with a blocked west doorway. The Roman stones are bedded vertically in the south-west quoin (against which a south aisle has been built) and their orientation has been set to match the orientation in the north-west quoin (also now with an adjoining later north aisle). .............................................................................................................................................................................. 21 Figure 2.26 The lower part of the flint rubble north-west nave quoin at Coldred church (TR 274 475) in Kent. ......... 22 Figure 2.27 The north-east nave quoin of St Edmund, Kingsdown (TQ 579 633). ...................................................... 22 Figure 2.28 The top of the flint rubble north-east nave quoin of St Giles, Kingston (TR 198 512) in Kent. ................ 22 Figure 2.29 The south face of Sompting church tower (TQ 161 056). On examination, the stone lithologies clearly reveal that above the first stage the quoin and pilaster-strip stones have been replaced with Caen Stone (see Potter, 2007b, for precise particulars). .................................................................................................................... 23 Figure 2.30 A classification of quoin stones according to their shape, proposed by Gilbert (1946). His term ‘clasping’ referred to a stone which was of no great vertical thickness and square in horizontal cross section (which without removing the stone from the wall is impossible to assess). ................................................................... 23 Figure 2.31 St Mary, Stow (SK 882 819), in Lincolnshire, is here viewed from the south-west. The south-west quoin of the south transept illustrates the type of quoining of ‘megalithic stones set alternately north-south and east-west’ by Brown (1925), which he recognised as being of Anglo-Saxon craftsmanship.......................................... 24 Figure 2.32 The lowest stones in the south-west nave quoin of St Mildred with St Mary de Castro (TR 145 575), Canterbury. The quoin was described by Brown (1925) as a ‘big stone quoin’ of Anglo-Saxon origin. The stones are of re-used Roman Marquise oolite from France and were described originally by Hussey (1858) as of megalithic character. Potter (2006b) describes the quoins of this church in some detail. The bedding of the lowest two stones is orientated vertically (visible on the second stone). ................................................................................... 24 Figure 2.33 Brown (1925) designated All Saints, Rockland (TL 994 960) as his exemplar of Anglo-Saxon long and short work, exhibited here in the north-east quoin of the nave. .............................................................................. 24 Figure 2.34 Variants in Anglo-Saxon quoin structure (partly after Taylor and Taylor, 1965, Figure 4): a. Megalithic side-alternate quoins set in rubble walling with each quoin stone equating to several courses of wall stone. This is probably the most common variety of AngloSaxon quoin. Typically, the walling sits upon a plinth of square section with the plinth stones set BH. viii

b. Face-alternate quoins emplaced in much the same style as ‘a.’ above. c. Randomly placed megalithic quoin stones. d. and e. Long and short quoins of two different styles. In d. the stones are cut to shape before they are built into the wall: in e. all the stones are cut back; in most instances this being undertaken after the quoin was built. f. A quoin built of materials like cobbles and Roman tiles. .......................................................................................... 24 Figures 2.35 to 2.40 Typical variations in Anglo-Saxon quoin styles: Figure 2.35 The north-west tower quoin of St Bene’t, Cambrige (TL 449 583), closely follows the long and short style of Figure 2.34e. Figure 2.36 The church of Barnetby-le-Wold (TA 061 090) exhibits in its south-east nave quoin stones set both in side- and face- alternate pattern (Figure 2.34a. and b.). Figure 2.37 The megalithic stones in the north-west nave quoin of Duntisbourne Rous church (SO 985 060) most closely relate to Figure 2.34a. Figure 2.38 Gosbeck church (TM 150 556), in Suffolk, in its north-west nave quoin displays smaller stones largely set in a long and short, but unlike Figure 2.34d. or e., in style. Figure 2.39 The unusual long and short style exhibited in the north-west quoin of Kirkdale church (SE 677 857), in Yorkshire. A tomb slab is inserted in the fabric of the west nave wall. Figure 2.40 The rubble and Roman tile corner to the south-east of the nave of St Peter, Iver (TQ 040 812) possesses no vertically orientated stones (Figure 2.34f.). ............................................................................................... 26 Figure 2.41 Using a similar display of quoin stones as shown in Figure 2.30, the stones are annotated with a notation proposed and used by the present author (Potter, 2005b). The notation (BVFR-BH-BVFL) refers to the bedding orientation within each stone and not to the stone shape................................................................................... 27 Figure 2.42 Typical stone settings for a pilaster-strip and the notation proposed and used by the present author (Potter, 2005b). This notation (BVFB-BH-BVEB) permits the bedding orientation of the stones to be distinguished. .................................................................................................................................................................. 28 Figure 2.43 The south face of Earls Barton church tower (SP 852 638) exhibits an excellent array of pilasterstrips. .............................................................................................................................................................................. 28 Figure 2.44 Typical stone settings for arch jambs together with the notation proposed and used by the present author (Potter, 2005b). This notation (BVFIA-BH-BVEIA) permits the bedding orientation of the stones to be distinguished. Anglo-Saxon stonemasons appear to have used only the orientations of stone of Bedding Vertical Face Into the Arch (BVFIA) and Bedding Horizontal (BH). Norman masons typically used Bedding Vertical Edge Into the Arch (BVEIA) and (BH). ......................................................................................................................... 29 Figure 2.45 Bedding stratification in sedimentary rocks and the resulting typical shape of hewn rocks (after Potter, 2005b).................................................................................................................................................................. 30 Figure 2.46 The church of St Thomas, Winchelsea (TQ 904 174), here viewed from the south-west, was inspired by King Edward I but probably never completed. The nave is absent and only the eastern walls of the transepts remain. In the transept walls, remnants of Viviparus limestone columns are badly weathered (see Potter, 2005d). ...... 30 Figure 2.47 Using the ‘grain’ of a rock to extract an elongated, uniform width stone, suitable for a pilaster strip. ‘A’ chisel entry to split stone along bedding plane; ‘B’ second chisel entry to break stone further. .............................. 31 Figure 2.48 Detail of one of the pilaster-strips in the south face of Earls Barton church (Figure 2.43). Note that both of the ‘long’ stones are placed BVEB. ................................................................................................................... 32 Figure 2.49 A typical Anglo-Saxon pilaster-strip in which the stones have been cut back to improve the decorative appearance, annotated with the terms proposed and used by the present author (Potter, 2005b; 2007c) to describe the pilaster. The amount of cut back (on the right of the pilaster) of overlarge stones, has been determined by the proud face width of the narrowest stone (in this instance the lowest stone shown). ......................... 32 Figure 2.50 Rock bedding planes and cut backs (after Potter, 2006c). Three ashlar stones of identical shape are shown; their bedding planes (BP) and the traces of these have been illustrated in the three different orientations that are possible. The potential success in trying to create a clean vertical arris and a smooth cut back (by chiselling in the direction of the arrows) will differ in each case: a. Will produce the best result. Having chiselled into the rock to the depth required the cut back should ‘fall away’, parting along its bedding planes. ix

b. Chiselling will tend to weaken the rock and the cut back portion will be difficult to remove. c. A complicated chiselling task and difficult to achieve smooth cut back surfaces. ................................................... 32 Figure 2.51 The Anglo-Saxon quoins in the church at Wittering (TF 056 020) in Northamptonshire are well preserved. The north-east chancel quoin, of long and short style, now abutted by an extension to the north (right) side, exhibits its east face. With the exception of the lowest (BVFR) stone above the plinth, the higher stones are cut back on their south (left) side. The cut back of the higher stones is in keeping with the lowest stone width. ........ 33 Figure 2.52 Elevations and cross sections of pilaster-strips and quoins, ‘without’ cut backs, in long and short work (partly after Potter, 2006c): a. Elevation of a pilaster-strip in which the ‘long’ stones have been prepared to matching width prior to insertion in the wall. The ‘short’ stones are of variable length (compare with Figure 2.42 in which all stones have been pre-cut). b. Cross section of ‘a.’ along line A-B, with all stones set flush to the wall surface. c. Cross section of ‘a.’ along line A-B, with pilaster face set forward of the wall. Note that in this instance the ‘short’ stones would need to be cut back to match. d. Elevation of a quoin (certain elements of perspective purposefully ignored) in which the ‘long’ stones have been prepared to matching width prior to insertion in the wall. Again the ‘short’ stones are of variable length. e. Cross section of ‘d.’ along line C-D, with all stones set flush to the wall surfaces. f. Cross section of ‘d.’ along line C-D, with the quoin faces set forward of the walls. In this rare instance the ‘short’ stones would have required to have been cut back. ................................................... 34 Figure 2.53 Elevations and cross sections of pilaster-strips and quoins, with cut backs, in long and short work (partly after Potter, 2006c): a. Elevation of a pilaster-strip in which the stones required cutting back to match in width. The width of the cut back has been determined by the proud face width of the narrowest stone (the lowest shown). Note the second ‘long’ stone could be described as ‘pear-shaped’. b. Cross section of ‘a.’ along line A-B. Removed cut back stone stippled. c. Elevation of a quoin (certain elements of perspective purposefully ignored) in which the stones required cutting back for stones to match in width. The width of the cut back has been determined by the narrowest points in the column (marked ‘S’). d. Cross section of ‘c.’ along line C-D. Removed cut back stone stippled. ................................................................. 35 Figure 2.54 The remnant pilaster-strip on the north nave wall of St Mary, Breamore (SU 153 188). This pilaster was selected by Brown (1903) to be illustrative of the presence of cut backs, although he failed to observe that the ‘long’ stones were cut back. Note that the lower ‘short’ stone has traces only of a possible cut back and that the strip width (proud face) relates to the width of the lowest visible stone which is not cut back (Pen for scale, 140mm.). ........................................................................................................................................................................ 36 Figure 2.55 All Saints church, Little Somborne (SU 382 326) is here viewed from the east. Details of this church are provided in Potter (2006a). The chancel arch is visible on the east wall, the chancel having been removed. A single pilaster-strip may be observed at the further (west) end of the nave north wall. ................................................. 37 Figure 2.56 The inside of the north nave wall of St Peter, Diddlebury (SO 508 853) in which a double-splayed Anglo-Saxon window is built. The wall exhibits well executed Anglo-Saxon herringbone work, the stones notably lining up with those in the window. Taylor and Taylor (1965, 212-213) provide a full explanation. ............... 37 Figure 2.57 Marton church (SK 840 817), in Lincolnshire, possesses Anglo-Saxon stones with vertically orientated bedding in a quoin in both the lower parts of the tower and nave (Potter, 2005b). The south wall of the lowest portion of the tower, as well as the remaining part of the west end of the nave, is generally thought to display Anglo-Saxon herringbone work (see Taylor and Taylor 1965, 412-414). Stocker and Everson (2006), however, judge all the work to be of Post-Conquest age, whilst accepting (p. 46) that ‘The tower…was attached to a church that was set within an already well-established graveyard’. ....................................................................... 37 Figure 2.58 St James the Less, Little Tey (TL 892 237), displays a band of ferruginously-cemented gravel in both its south chancel wall and apse. This band is set in walls of flint cobbles and is disrupted where later windows have been inserted. Potter (2001a) regarded the church as being of late Anglo-Saxon age. .......................... 38

x

Figure 2.59 St Andrew, Heybridge (TL 855 081), like Little Tey, displays a band of ferruginously-cemented gravel, on this occasion high in the first stage of its early tower. A later window again disrupts the band. ................. 39 Figure 2.60 Ornamentation, provided by ferruginously-cemented gravel, in the round, west, flint rubble tower of St Lawrence, Beeston (TG 328 219). The band continues in the original part of the west nave wall (see also Figure 2.61). ................................................................................................................................................................... 39 Figure 2.61 The north nave wall of Beeston Anglo-Saxon church (Taylor and Taylor, 1965, 61) also exhibits its originality. The tower band continues along the nave wall as far as the original north-east nave quoin which was also constructed of ferruginously-cemented gravel. ....................................................................................................... 39 Figure 2.62 The church of St Martin, Barholm (TF 090 110) preserves in its south nave wall an Anglo-Saxon blocked doorway in which the stones in the jambs are placed in appropriate vertically bedded style (Potter, 2005b, 209). A pilaster rises above the doorway. To the west, within the porch, is the currently used Norman doorway. Within the Anglo-Saxon work different contrasting Middle Jurassic limestones have been used to provide ornamentation. ................................................................................................................................................. 39 Figure 2.63 St Botolph church in Hadstock (TL 558 447) has been described in detail by Rodwell (1976), and this south wall of the church nave by Potter (2008d). A band of Bunter cobbles runs across the nave walls, with its base at the level of two Anglo-Saxon double-splayed windows. Note that the band may be traced into the outer splay of the windows. .......................................................................................................................................... 40 Figure 2.64 Diagrammatic sketch of an Anglo-Saxon or ‘Patterned’ wall illustrating the use of stones orientated with their bedding vertical. FB = Face bedded, EB = Edge bedded, BH = Bedded horizontally. In the doorway set in the wall, BVFIA = Bedding vertical face into the arch. ........................................................................................ 40 Figure 2.65 The south nave wall of Escomb church (NZ 189 301), in Durham, is constructed of blocks of Carboniferous sandstone. These blocks are set in a ‘Patterned’ style of FB-EB-BH stones. Regrettably, largely because of surface grime on the stones, it proves impossible to photograph the various stone bedding orientations. .................................................................................................................................................................... 41 Figure 2.66 The south wall of the chancel of Jarrow church (NZ 339 652), also exhibits stones emplaced in patterns of FB-EB-BH stones. Despite low sunlight the photograph fails to accurately portray the patterning. .......... 41 Figure 2.67 The lower part of the west face of the tower at Cabourne (TA 139 018), in north Lincolnshire. The upper part of the tower was restored and rebuilt in 1872. Although the western corners of the lower component of the tower are constructed with a number of stones placed in vertically bedded orientation (in Anglo-Saxon fashion), the western doorway jambs are built entirely with horizontally bedded (BH) stones. Taylor and Taylor (1965, 127) regarded the doorway, on some of its architectural characteristics, as of late Anglo-Saxon age. The arch includes a tympanum set BVEIA in Norman fashion and Stocker and Evison (2006) placed the whole of this lower portion of the tower in that period (see Potter, 2005b, 198). ........................................................................ 42 Figure 2.68 The church of St Laurence, Bradford on Avon (ST 824 609) presents building work of a variety of periods. The south wall of the chancel illustrated in this figure shows a double-splayed window which from its BH stonework shows no Anglo-Saxon characteristics and has been completely reconstructed. The south-east chancel quoin stones are set in typical Anglo-Saxon style and a number have large cut backs. .................................... 44 Figure 2.69 The church at Brixworth (SP 747 713) was erected in a region of England where quality building stone is exceptionally common, and yet, in those structures where Anglo-Saxon masons might have been expected to use such favourable stone (the tower’s south doorway, the south-east nave quoin, etc.) it is absent (see Potter, 2009b). The church is viewed here from the south-west. ............................................................................ 45 Figure 2.70 The apsidal chancel of Wing church (SP 880 225) viewed from the south-east. The apse possesses many unusual features that are atypical of Anglo-Saxon work. The pilaster stones are, for instance, all set BH. ........ 46 Figure 2.71 A distribution map of Anglo-Saxon long and short quoins as prepared by Taylor (1978, 945) and reproduced by kind permission from that publication. The figure has been modified by the present author as follows: localities underlined have these quoins created in Middle Jurassic oolitic limestone; localities with broken underline have the quoins built of Quarr Stone. .................................................................................................. 47 Figure 2.72 The church of St Mary de Castro, Dover (TR 326 417) is here viewed from the south to show the adjacent Pharos and Dover Castle (left). ....................................................................................................................... 49 xi

Figure 2.73 Detail of the north-west quoin of the nave of St Mary de Castro, Dover. Although this part of the quoin is built in ‘Patterned’ style, the stone is of a re-used Roman origin. The large white stone is Marquise Oolite from northern France. See Potter (2006b) for a full explanation. ...................................................................... 49 Figure 2.74 This distribution map showing the concentration of Anglo-Saxon double-splayed windows in south and south-east England was prepared by Taylor (1978, 841). Localities underlined involve an element of dressed stone in the window(s) concerned. Permission to reproduce this figure is gratefully acknowledged. ............................ 50 Figure 2.75 The distribution map of hoodmouldings displayed in Anglo-Saxon churches as compiled by Taylor (1978, 931), from which the map has been reproduced. In the south-east of England these features are absent. The majority of the localities relate to the proximity of the Middle Jurassic limestones. ............................................... 51 Figure 2.76 This distribution map of the presence of churches of Anglo-Saxon origin with eastern apses in England was also compiled by Taylor (1978, 1029). Only the small east church at Hexham lies north of the Wash. Taylor (1978, 1028) makes the point that the apsed churches extend ‘widely beyond the early kingdom of Kent’. The evidence is strongly in favour of their origin being related to the influence of the regional geology, and to areas where stone in the form of rubble predominates. ........................................................................................ 52 Figure 2.77 Much of the south wall of the chancel of the early church of St Martin and St Paul, Canterbury (TR 158 575) is displayed in this figure. The doorway with the heavy lintel of very glauconite-rich Hythe Beds sandstone was almost certainly originally quarried by the Romans from close to Hythe. It is believed to have served as an entrance to a south porticus. The round headed doorway constructed of Roman tiles is regarded as a later Anglo-Saxon insertion. The south-east nave quoin (left) is also constructed of re-used Roman material, the white stone blocks being of Ditrupa limestone (Calcaire Grossiere) from France. ....................................................... 54 Figure 2.78 A typical Norman ‘Romanesque’, bonded ashlar wall with a broad shallow pilaster and a clasping quoin. Where they might have overlapped the proud faces, stones will have been, of necessity, cut back. The vast majority of the stones in the wall and the structures will have been placed with their bedding orientation horizontal, but occasionally where a shorter stone is required it may be cut and have an edge bedded setting (after Potter, 2006c). ....................................................................................................................................................... 55 Figure 2.79 The south wall of the nave of Dymock church (SO 700 312), showing a Norman pilaster in sandstone of Lower Devonian age (Lower Old Red Sandstone). All the stones are set BH and where they would have extended beyond the pilaster width they have been cut back. .............................................................................. 55 Figure 2.80 The south wall of the nave at St Nicholas, Worth (TQ 302 362). In the upper portion of the southwest quoin, the proud face of the quoin suddenly changes in width: at this point the quoin and adjoining wall change from Anglo-Saxon to Post-Conquest in fashion and the cut back stones are of different style (see Potter, 2006c for detail). ............................................................................................................................................................ 55 Figure 3.1 Strathpeffer Pump Room (NH 486 583) was built as recently as 1892 and the walls were constructed of rock which is presumed to be dark, Devonian, Middle Old Red Sandstone. The stones of the quoins were placed in side-alternate style with the bedding orientated vertically in each stone (BVFR or BVFL). This style of construction was followed by some masons in Scotland, particularly in the 1730 to 1870 period, and was virtually a copy of the much earlier ‘Patterned’ style (see section 3.3.2). BH tie stones, however, tend to be absent in the more recent style. Note that the stones have been both cut and dressed. .................................................. 58 Figure 3.2 The south-west quoin of St Serf, Kirkton, Burntisland (NT 232 864), is built of relatively large blocks of Lower Carboniferous sandstone set in side-alternate fashion. Many of the stones are placed with their bedding orientated vertically in ‘Patterned’ style. From the ground upwards the orientations of most stones can be read with a hand lens as follows: stone 1, ?; BVFL; stone 3, ?; BH; BH; BVFL; BVFR; BH. ............................................ 58 Figure 3.3 The south-west quoin of St Mary, Rothesay, Bute (NS 085 636). This quoin has been rebuilt at the base but from stone 5 and above (stone 9 has been replaced) the orientations in the Upper Devonian sandstone blocks exhibit a ‘Patterned’ style. .................................................................................................................................. 59 Figure 3.4 The ruined church of Little Dalton (NY 090 747) was briefly described by the Royal Commission for Historical Monuments for Scotland (RCAHMS, 1920, 42) who described the church as ‘probably early 16th century’. The church is here viewed from the south-east and the east wall (right) is clearly of different and much larger stone composition than the south wall. Large blocks of Permian create the south-east quoin, which like the north-east quoin, contains stones set in ‘Patterned’ style. ............................................................................................... 59 xii

Figure 3.5 Old St Andrew church, Gullane (NT 480 827) provides a complex, much altered and ruined church site. The south doorway to the original nave, although somewhat modified, still shows stones of Lower Carboniferous sandstone exhibiting BVFIA orientation in ‘Patterned’ style; here shown in the east jamb. ................. 60 Figure 3.6 The small, single-celled ruined church at Glenearn (NO 107 164) is built of Upper Devonian sandstone blocks. Each of the four principal quoins displays vertically orientated stones set in ‘Patterned’ style. The view is towards the north-east quoin; the east wall stands on a rough square plinth, and the church appears to stand on early foundations. ............................................................................................................................................ 60 Figure 3.7 The ruined church of St Columba, Uidh, Eye peninsula, Lewis (NB 485 323), viewed from the south-east. The original church was apparently single-celled, but a large slightly narrower west annexe has been added (it has been suggested in the late 15th or early 16th century; RCAHMS, 1928, 13). The original church is built of local Lewisian gneiss beach boulders packed with small broken fragments of gneiss, extensive repairs having been completed with Permian, red sandstone, from the Stornaway ‘Beds’. The eastern quoins both include a number of blocks of gneiss which exhibit vertical orientation. ...................................................................... 60 Figure 3.8 On the south wall of St Columba, Uidh, a 13th-14th century doorway remains. To its left (west), the west jamb of an earlier infilled doorway is visible, this doorway has some evidence of ‘Patterned’ workmanship. ..... 61 Figure 3.9 The early Christian cross at Kildalton church, Islay (NR 458 507), viewed from the south-west. Although modified, the four principal quoins at Kildalton, in their lowest stones, show some evidence of ‘Patterned’ emplacement. .............................................................................................................................................. 61 Figure 3.10 This boulder in the north-west wall at Kildalton church shows Pholas borings, indicating its collection from the sea shore. ....................................................................................................................................... 61 Figure 3.11 The site of the priory church of St Peter, Restenneth (NO 482 516) viewed from the north-west. The roof line of the nave is clearly marked on the tower. The quoins of the tower are set in ‘Patterned’ style (Potter, 2008c). ........................................................................................................................................................................... 62 Figure 3.12 The east side of the tower arch at Restenneth church, the jamb stones of which, although now somewhat modified, show evidence of ‘Patterned’ insertion. ........................................................................................ 62 Figure 3.13 St Rule, St Andrew’s (NO 515 167) is constructed of a Lower Carboniferous fine grained sandstone. The church has recently been described by Potter (2008c) and is here viewed from the north-west. ............................ 63 Figure 3.14 The south-west quoin of the chancel of St Rule, St Andrew’s is set with its stones in ‘Patterned’ style. From the plinth the stones read upwards: BH, BVFL, BVFR, BVFL, BVFR. In the first four courses, the clearly visible stones in the south wall are, with the exception of BVFL quoin stones, face bedded. ........................... 63 Figure 3.15 The western tower arch of St Rule, St Andrew’s was rebuilt and subsequently blocked. ......................... 63 Figure 3.16 The south wall of the chapel of St Margaret, Edinburgh Castle (NT 253 735) reveals evidence of four styles of building at different levels, one of which may represent ‘Patterned’ colour banding (see Potter, 2008c). ........................................................................................................................................................................... 63 Figure 3.17 Detail of stones which are believed to have been inserted in ‘Patterned’ style in the north-west quoin of the Edinburgh Castle chapel. ................................................................................................................................... 64 Figure 3.18 Abernethy round tower (NO 190 165) is here viewed from the north. The change in stone type in the tower (referred to in section 3.3.4) may be observed to the right of the doorway. ........................................................ 64 Figure 3.19 The doorway to Abernethy round tower, in which all the stones were originally placed in ‘Patterned’ style. .......................................................................................................................................................... 65 Figure 3.20 The illustrations of Abernethy round tower as produced by Brash (1857-60, Plate XXXII). The doorway jambs as drawn should be compared with those in the recent photograph. ...................................................... 66 Figure 3.21 The doorway to the round tower at Brechin Cathedral (NO 596 601) possesses all the characteristics of being constructed in ‘Patterned’ style. The stones in the doorway are set BVFIA and exhibit cut backs. ................ 66

xiii

Figure 3.22 The illustrations to the paper by Jervise (1857-60) of the round tower at Brechin. There are errors in the drawing of the doorway............................................................................................................................................. 66 Figure 3.23 A plan of St Magnus Church, Egilsay. Key: A1 – south doorway; A2 – north blocked doorway; B1, B2 – early single-splay blocked windows to nave; C1, C2 – early single-splay blocked windows to chancel; D – access doorway to tower (with nearly identical doorway on first floor above); E – above chancel arch, doorway nave first floor to chancel first floor, with step; F- first floor early window to tower; G – later rectangular window to chancel first floor; H1 – late rectangular window above south nave doorway (preserves wooden frame); H2 – late blocked rectangular window (preserves part of wooden frame); H3 – late blocked and altered, rectangular window ground floor tower; J – trace of original narrower chancel arch. ................................................... 67 Figure 3.24 The problematical church of Egilsay, Orkney (HY 466 304) as viewed from the south-east. .................. 68 Figure 3.25 The nave and tower of Egilsay church viewed from the south and drawn by Meyer for publication in Dietrichson and Meyer (1906). .................................................................................................................................. 69 Figure 3.26 The crow-stepped or corbie gable at the east end of the nave of Egisay church viewed from the south-west. .................................................................................................................................................................... 70 Figure 3.27 The corbie gable at the west end of the nave of Egilsay church, viewed from the south. The small rectangular window is referred to in the text. ................................................................................................................ 70 Figure 3.28 A view of the interior of the west end of the Egilsay nave to show the relationship with the tower and the access entrances to the tower at ground and first floor levels. .......................................................................... 71 Figure 3.29 A view of the interior of the east end of the nave of Egilsay church, showing the chancel arch, the simple doorway, with step, at first floor level and, the corbie or crow-stepped gable. ............................................... 71 Figure 3.30 The blocked north doorway in the north nave wall of Egilsay church. ...................................................... 71 Figure 3.31 The blocked north window in the north wall of Egilsay church has been infilled in the same style as the doorway. ................................................................................................................................................................... 71 Figure 3.32 Each of the jambs of the chancel arch show evidence of having been partially destroyed to widen the arch. The south jamb is here viewed from the north-west. It can be noted that the base of the jamb projects north of the chancel wall. ............................................................................................................................................... 72 Figure 3.33 View of the nave and tower of Egilsay church from the north. The fabric of large face bedded stones and smaller horizontally set rubble should be noted. .................................................................................................... 73 Figure 3.34 The south nave doorway of Egilsay church is here viewed from the south-east. The style of the stonework of the wall noticeably changes about a metre from the ground to include large face-bedded stones. ......... 73 Figure 3.35 The south-east chancel quoin of Egilsay church has many stones set to the BVFR-BVFL ‘Patterned’ style. These stones are often interspersed with smaller BH stones. Large stones 7 to 9 are visible, see text for detail. ............................................................................................................................................................................ 73 Figure 3.36 The south-west nave quoin of Egilsay church. In this quoin the stones are placed as follows: stones 1-6 BH, 7 BVFR, 8-13 BH, 14 BVFL, 15-17 BH, 18 BVFR, 19-24 BH, etc. The larger stones 7, 14 and 18 can be distinguished. (The pen is 135mm. in length). ......................................................................................................... 74 Figure 3.37 Stones below the infilled window in the south chancel wall of Egilsay church are clearly face bedded and one stone preserves fossil ripple marks. ...................................................................................................... 74 Figure 3.38 The keeill, or small early chapel, known as Lag ny Keeilley (SC 216 746), which nestles in the cliffs near Niarbyl Bay, Isle of Man, serves as a local picnic site. Better preserved than most keeills its walls are estimated to only be of the order of 675mm. thick. ....................................................................................................... 75 Figure 3.39 One of many early Christian Celtic crosses in the Isle of Man. This, partly repaired cross occurs in the churchyard of St Adamnan, Lonan (SC 426 794). Apart from an early plinth and large worn boulders at the base of certain walls no trace of an early church is really present at the church site. .................................................. 75 xiv

Figure 3.40 These three stones, set BVFL, BH, BVFL, occur as part of a short quoin at the base of the north-east corner of the north transept of the Cathedral of St German, Peel (SC 242 846). The cathedral was not built until the 13th century and it is difficult to relate either the quoin or its short length of adjoining wall to the cathedral or other adjacent earlier buildings (lens cap 50mm in diameter). ...................................................................................... 76 Figure 3.41 The ruined church of St Patrick, St Patrick’s Isle, Peel (SC 242 847) viewed from the south-west. A summoner’s oratory or pulpit stands on the west side of the nave.................................................................................. 76 Figure 3.42 Detail of the south wall of the nave of the church of St Patrick, Peel. The herringbone work in the wall succeeds several courses of Peel sandstone blocks laid in ‘Patterned’ style. See text for detail. ........................... 77 Figure 3.43 Viewed here from the east, the detached round tower to the west of the church of St Patrick, Peel, shows considerable evidence of alteration. ................................................................................................................... 77 Figure 4.1 A photographic copy of the 1:500,000 scale, Bedrock Geological Map of Ireland, produced by the Geological Survey of Ireland (2006) and reproduced by kind permission. The key to the different geological formations is too complex to display here, but see Table 4.1. ....................................................................................... 80 Figure 4.2 A nomenclature for the possible dispositions of stones which might be placed in an anta. ......................... 82 Figure 4.3 The larger, and more modern, ruined church at Coole, Co. Cork (W 860 952) is viewed here from the south-west. The church, largely built of Upper Devonian sandstones, originally consisted of a single chamber, which today forms the ruined nave, to this the chancel was probably added in the 13th century. The present nave originally possessed antae at each corner, of these only that at the north-west corner now clearly remains. The stones of the remaining anta, with one exception, are all laid BH and are thought to be of post-‘Patterned ‘ style in date. At some date subsequent to the loss of the south-west anta, the quoin (and elsewhere) was rebuilt, this in a much later but similar style to that practiced in the ‘Patterned’ period. ...................................................................... 83 Figure 4.4 Map of Ireland indicating the positions of the churches described in sections 4.5.2 to 4.5.39. .................. 85 Figure 4.5 A simple geological map of Ireland to show its principal geological structures. This map is reproduced from Whittow (1974) by the kind permission of the publishers, Penguin Books......................................... 86 Figure 4.6 The west gable and the adjoining round tower to the ruined Ardmore Cathedral (X 188 774). The remarkable arcading and ‘Romanesque’ sculpture on the west gable has been extensively referred to in the appropriate literature. The lack of symmetry to the arcading within the wall should be noted. ................................... 87 Figure 4.7 Due principally to grave build up, viewed from the south only the upper portions of the walls of Ardmore Cathedral are evident. Note the string course terminates unusually near the top of the south doorway in the nave wall. ................................................................................................................................................................. 87 Figure 4.8 Long observed changes in the style of stonework are evident in the north wall of the present chancel of Ardmore Cathedral. Most authors have regarded the larger stones in the lower part of the wall as being of ‘pre-Romanesque’ age. .................................................................................................................................................. 87 Figure 4.9 The north wall of the present chancel to Ardmore Cathedral viewed from the inside shows different styles of stonework. The blocked window was once clearly inserted into an earlier wall. Note that this wall has also been extended towards the east. ............................................................................................................................. 88 Figure 4.10 Viewed from the north-west, the north-east quoin stones of Ardmore Cathedral nave can be compared with the early lower chancel wall. At the east end of this wall (between the two blocked windows) there is a further fabric change, representing a chancel eastward extension. ............................................................... 88 Figure 4.11 The north-east quoin of Ardmore Cathedral is here viewed from the north-east. The fabric in the adjoining east and north walls is unlike the stonework in the quoin and the stones were probably, at least in part, re-used from the western quoins of the earlier church. .................................................................................................. 88 Figure 4.12 The north-west anta of Loorha church (M 920 046) is viewed here from the south-west. The anta was in the process of repair and newly pointed. The stone bedding orientations are provided in the text. .................... 89 Figure 4.13 The lower 1.2m. of the south-west anta at Loorha has been rebuilt; higher, although repaired, the rocks are largely set to the ‘Patterned’ style. .................................................................................................................. 89 xv

Figure 4.14 The ‘Romanesque’ chancel arch at Monaincha Abbey (S 170 884) viewed here from the nave. Note that the relieving arch is not fully continuous and that there is a change in wall fabric style approximately half way up the gable. .......................................................................................................................................................... 90 Figure 4.15 The ‘island’ setting to Monaincha Abbey. This view from the north-west shows both the sacristy (left) and the high cross. ................................................................................................................................................ 90 Figure 4.16 The chancel of Monaincha Abbey viewed from the south-east to show the repaired pilasters at the quoins. ........................................................................................................................................................................... 91 Figure 4.17 The south wall of the chancel at Monaincha and the half-column pilaster near the angle with the nave. The window in the south wall is designed to hold a shutter. ............................................................................... 91 Figure 4.18 The ‘Romanesque’ west doorway at Monaincha Abbey viewed from the west. ...................................... 92 Figure 4.19 The same west doorway at Monaincha viewed from the east and from inside the nave. There is some visible evidence of repairs or alteration. .............................................................................................................. 92 Figure 4.20 Detail of the south jamb of the west doorway to Monaincha Abbey as viewed from the east. Where the outer order of the hoodmould extends into the architrave of the jamb the stones vary, the rounded stones are cut back; those with a rectangular section would appear to be replacements. .............................................................. 92 Figure 4.21 Further detail of the top of the south jamb of the west doorway at Monaincha. Note the cut back stones. ............................................................................................................................................................................ 92 Figure 4.22 Ardpatrick church (R 637 208), viewed from the north-west and across the ruined stump of the round tower. The extent of the ivy cover is well illustrated. ......................................................................................... 93 Figure 4.23 The exterior of the south wall of the nave of Ardpatrick church is terminated at its south-west quoin with ‘Patterned’ stone insertion. The stonework viewed here is quite unlike that shown in the north wall (Figures 4.24 and 4.25) and it is considered to be of the earlier ‘Patterned’ period...................................................................... 94 Figure 4.24 The stonework typical of the exterior of the north wall of the nave of Ardpatrick church. This wall was probably constructed in the 12th century. Compare with Figure 4.23. .................................................................... 94 Figure 4.25 The inside of the north wall of the nave of Ardpatrick church exhibits stonework much the same as the outside of the same wall (Figure 4.24), but unlike that of the earlier south wall (Figure 4.23). .............................. 94 Figure 4.26 Killulta church, Limerick (R 434 534) viewed from the south-west displays two distinct styles of stonework. That in the south wall (right) may well be largely original, the stonework in the south-west quoin and the west gable being, in comparison, relatively recent. ................................................................................................ 95 Figure 4.27 Detail of the north-east quoin of Kilrush church, Limerick (R 556 567). The lowest five stones in this quoin above the plinth are set BVFR, BH, BVFR, BVFL, BH in ‘Patterned’ style. .............................................. 95 Figure 4.28 The west wall of Kilrush church showing the flat-headed doorway. The south (right) jamb of this doorway has five small replacement stones in its lower portion, but higher through stones are placed BVFIA. .......... 96 Figure 4.29 The west face of St Molua church, originally situated on Friar’s Island, now rebuilt in Killaloe (R 701 728). The west doorway to the nave has an ‘Escomb style’ appearance. ................................................................ 97 Figure 4.30 The inside of the chancel roof to Teampul Molua viewed towards the east window. This roof has been extensively rebuilt. .............................................................................................................................................. 97 Figure 4.31 The small ruined church at Noughaval (R208 967) has been greatly modified and a barrel vaulted roof has been fitted between its gables. The much larger stones occurring at the church corners appear to be of early workmanship and the quoin stones tend to show ‘Patterned’ orientations. .......................................................... 98 Figure 4.32 Artistic licence replaces any real accuracy in this painted wash of the Oughtmama site by Cooke (1842-3). It shows the presence of the three churches, with presumably the west church to the left (west). The chancel is apparently only represented by one low wall and the alignment of each church has been altered. (Print kindly reproduced from a copy held by Clare County Library). ..................................................................................... 99 xvi

Figure 4.33 Oughtmama west church (M 304 078) viewed from the south-east reveals distinct masonry fabric differences between the chancel (right) and the largest area of the visible nave wall. ............................................. 100 Figure 4.34 The west door to the nave at Oughtmama west church (viewed here from the north-west) possesses jambs in typical ‘Patterned’ orientation. ...................................................................................................................... 100 Figure 4.35 The exterior of the west window in the south wall of Oughtmama west church has from the clear evidence of masonry disturbance been set into an earlier wall fabric. The window has distinct ‘Romanesque’ characteristics and the fitting for a shutter. .................................................................................................................... 101 Figure 4.36 The close proximity and alignment of the churches at Oughtmama is illustrated in this view, taken from the south-east, of the central and, more distant, west church. ............................................................................... 101 Figure 4.37 The small church of Temple Cronan (M 289 000) seen from the south-east. A house (or slab) tomb can be observed in the foreground. .............................................................................................................................. 102 Figure 4.38 Temple Cronan church is here viewed from the north-west and this corner is of larger stones. The blocked west doorway and its replacement north doorway are also visible. ................................................................. 102 Figure 4.39 The south-west quoin, the top of the west doorway and carved ‘Romanesque’ heads are all visible in this view of Temple Cronan church. ............................................................................................................................. 103 Figure 4.40 The interior view of the west doorway of Temple Cronan church. Compare with Figures 4.38 and 4.39. ............................................................................................................................................................................. 103 Figure 4.41 The north wall of Tuamgraney church (R 637 830) viewed from the north-east. Pre-‘Romanesque’ walling creates the wall fabric at the west (farther) end, ‘Romanesque’, possibly 12th century, walling at the east. ..... 104 Figure 4.42 The ‘Patterned’ style west doorway at Tuamgraney church viewed from the north-west. ...................... 104 Figure 4.43 Killeenemeer church (R 775 070) is viewed here from the south-west to display the west antae. ......... 105 Figure 4.44 Detail of the western end of the northern wall of Killeenemeer church. This portion of the wall is probably of ‘Patterned’ style. Of the three full courses shown the lower two are bedded in normal BH orientation. The third course has the stones inserted with the bedding placed BVFB. ................................................. 105 Figure 4.45 The west face of Labbamolaga small ruined church (R 764 176) showing the west ‘Patterned’ style doorway. ........................................................................................................................................................................ 106 Figure 4.46 The north-west anta to Labbamolaga small church. The folder resting against the structure is 310mm. long. Some of the stones in the anta can be seen in this figure to be emplaced BH, and those in the door jambs BVFIA. Compare with Figure 4.47. .................................................................................................................. 106 Figure 4.47 The south-east anta to Labbamolaga small church extends much further from the gable than those at the west end of the church (Figure 4.46). The folder is 310mm long. ........................................................................ 107 Figure 4.48 The modified east window at Kilgarvan church (W 013 734) viewed from the east. Steel rule length 325mm. .......................................................................................................................................................................... 108 Figure 4.49 Kilgarvan ruined church viewed from the north-east to show the north-east quoin. ................................ 108 Figure 4.50 The west doorway of Ratass church (Q 853 141) is built of Upper Devonian sandstone. The doorway has an architrave which has been created by cutting back the various BVFIA and BH stones which form the jambs. The whole structure is typical of work in a ‘Patterned’ style (rule 325mm. long). ............................. 109 Figure 4.51 The south-west nave corner of Ratass church showing the antae. Note that two of the highest stones in the south-west anta are visibly set BVFL and these are thought to be replacements. Most of the later replacement stones are in grey limestone as may be observed high in the west gable. ............................................... 109 Figure 4.52 The south-east quoin of the ‘Romanesque’ chancel of Ratass church which is principally created in sandstone. .................................................................................................................................................................... 110 xvii

Figure 4.53 The north-east quoin of Rattoo church (Q 878 336) viewed from the north. The lowest stones in the quoin are of calcareous sandstone, noticeably larger, and set in ‘Patterned’ style. ........................................................ 110 Figure 4.54 Detail of stones 2 and 3 in the north-east quoin, shown in Figure 4.53, of Rattoo church. The bedding orientation in these stones is clearly visible as BVFL, BVFR. Subsequent to work on the site in 2009, these stones may no longer be in the same position in this quoin. ................................................................................ 110 Figure 4.55 Viewed from the south-east the problematical church at Agha, Carlow (S 730 654) displays its two contrasting wall masonry styles. The well coursed and hammered style of granite boulders at the western end of the building would be considered of ‘Romanesque’ origin in other churches. ............................................................ 111 Figure 4.56 Agha ruined church is here viewed from the north-west. The two styles of masonry are well exhibited on the inside of the east gable. The inside of the east window is noticeably of ‘Romanesque’ character....... 112 Figure 4.57 From the north-east the ‘Romanesque’ style of the inner Agha church walls is well displayed. Note that above the modified ‘Patterned’ style reset doorway the interior wall structure is of rubble. ................................. 112 Figure 4.58 The interior of the south wall of Agha church here exhibits both the ‘Romanesque’ hammered granite boulders and the rubble interior wall infill (now cemented into position). ........................................................ 113 Figure 4.59 A general view of the Glendalough site (T 122 968) viewed from the east. ............................................. 114 Figure 4.60 Courses 1 to 8 of the north-west anta at Glendalough Cathedral, in which the phyllitic/schistose stone lineations can be observed (see text) and display a ‘Patterned’ style. .................................................................. 114 Figure 4.61 Courses 2 (top only) to 6 of the north-east anta at Glendalough Cathedral in which stone lineations show the ‘Patterned’ style with moderate clarity. ......................................................................................................... 115 Figure 4.62 The outside of the ‘Patterned’ style west doorway at Glendalough Cathedral (Folder 310mm. long). ...... 115 Figure 4.63 The inside of the same ‘Patterned’ doorway illustrated in Figure 4.62. The jambs of the frame have been cut into to provide a setting for a door at some post- ‘Patterned’ date. ............................................................. 116 Figure 4.64 The north-east portion of the north nave wall at Glendalough Cathedral. This illustrates the two masonry styles with the lower three ashlar courses emplaced in ‘Patterned’ style. .................................................... 116 Figure 4.65 This joggled joint occurs on the inside of the west wall of Glendalough Cathedral. (Rule marked in centimetres). ................................................................................................................................................................... 117 Figure 4.66 The BH capping stones at the top of the south-west anta at Glendalough Cathedral. ............................... 117 Figure 4.67 The ruined church of St Mary, Glendalough (T 122 968) viewed from the north. .................................. 118 Figure 4.68 The west doorway to the church of St Mary, Glendalough has its stones set in ‘Patterned’ style. The doorway is viewed here from the inside. ...................................................................................................................... 119 Figure 4.69 The outside of the west doorway at St Mary, Glendalough preserves its architrave the width of which has been determined by the north (right) jamb stones. ......................................................................................... 119 Figure 4.70 The cross which is carved on the underside of the granite lintel to the west doorway of St Mary, Glendalough. ............................................................................................................................................................... 119 Figure 4.71 The base of the west jamb of the north doorway to St Mary, Glendalough has been built of the same Middle Jurassic limestone as seen at Glendalough Cathedral. It is possibly Dundry Stone from near Bristol. ............ 120 Figure 4.72 Replacement stones of basaltic lava are difficult to distinguish from the phyllite/schist in this window in the south wall of the church of St Mary, Glendalough, viewed from the exterior. They make up most of the frame. ................................................................................................................................................................... 120 Figure 4.73 Temple Ciarán, Clonmacnoise (N 010 306) viewed from the south side, with the River Shannon and other ecclesiastical buildings in the distance. ................................................................................................................ 121 xviii

Figure 4.74 The west gable and doorway of the ruined Temple Ciarán. Looking ‘old’ it preserves little that is original fabric. ............................................................................................................................................................ 121 Figure 4.75 The east gable and antae of Temple Ciarán. The north-east anta (right) preserves elements of its ‘Patterned’ originality in its lower stones. Putlog holes are evident in the east wall. ................................................. 121 Figure 4.76 The south doorway of the Dulane church ruin (N 741 788) viewed from the north (interior). Note the wall fabric which is quite unlike that in the west wall (Figure 4.77). The wall in which this doorway occurs has probably been rebuilt. ............................................................................................................................................ 122 Figure 4.77 The west wall of Dulane church. The west, like the south doorway, cannot be seen to its full height because soil covers the threshold to some depth. ......................................................................................................... 123 Figure 4.78 The church of St Patrick, Duleek (O 045 684) is here viewed from the south. There is today a significant fall in land to the west (and north). ............................................................................................................. 124 Figure 4.79 A remnant north-east quoin and the visible north plinth below the north wall at Duleek church ............. 124 Figure 4.80 St Columcelle’s ‘House’, Kells (N 740 758) is here viewed from the south-west. The wall fabric change near the base of the south wall can be distinguished. The south door is reasonably modern and now provides the access to the building. ................................................................................................................................ 125 Figure 4.81 A closer view of the south wall of the Kells building reveals more clearly the nature of the older wall fabric which includes river derived cobbles and boulders of Carboniferous limestone and sandstone. ............... 125 Figure 4.82 The interior of the north wall of the Kells building also shows the same fabric change. The lower stones are spaced with mortar and less well coursed. .................................................................................................. 126 Figure 4.83 The ‘propping arches’ which are built into the interior of the roof of St Columcelle, Kells. ................... 126 Figure 4.84 The interior of the west wall at Kells displays a fabric change at gable level, indicating rebuilding at the time of construction of the vault and roof. .............................................................................................................. 126 Figure 4.85 A view of the Fore monastic site (N 510 704) from the south-west, with St Feichin church in the foreground. ................................................................................................................................................................... 127 Figure 4.86 The west doorway of St Feichin church, Fore is constructed in ‘Patterned’ style. Note that there is a change in wall fabric to the south (right) of, and above, the doorway. .................................................................... 128 Figure 4.87 The west face of St Mel, Ardagh (N 204 686) and the north-west anta. .................................................. 128 Figure 4.88 The south-west anta at Ardagh church displays two courses of ‘Patterned’ stone laid BH and then BVFL, BVFR. Stones at level three are both rebuilt and joggled. (Rule in centimetres). ............................................. 129 Figure 4.89 The east wall of St Mel, Ardagh showing the megalithic blocks of face-bedded Lower Carboniferous calcareous sandstone capped more recently with limestone. The calcareous sandstone south-east anta is also shown with the nearest and lowest stone placed BVFR. The plinth is of flaggy slightly calcareous sandstone. ...................................................................................................................................................................... 129 Figure 4.90 Drumacoo church (M 396 168) preserves its ‘Patterned’ doorway which is here viewed from the north-west. ................................................................................................................................................................... 130 Figure 4.91 The west doorway of Drumacoo church as seen from the interior. ......................................................... 130 Figure 4.92 The end of the north wall of the original stone Drumacoo church and its north-east quoin is marked by two aligned stones and the wall fabric change. ......................................................................................................... 131 Figure 4.93 Dune sands cover much of St Cavan church, Inisheer (L 986 025). The church is here seen from the south-east. .................................................................................................................................................................... 132 Figure 4.94 The west doorway of St Cavan, Inisheer, observed from the east (inside) of the nave. The massive lintel stone of the doorway stands proud and the jambs have been modified. ........................................................... 132 xix

Figure 4.95 The west doorway viewed in Figure 4.94 is here seen from the west, the exterior. ................................ 133 Figure 4.96 The exterior of the window in the east gable wall of St Cavan, Inisheer, shows distinct ‘Patterned’ characteristics as may be observed from the stone bedding orientations in the jambs. ............................................... 133 Figure 4.97 The triangular-headed window in the south-east of the chancel at St Cavan, Inisheer, internally possesses all the characteristics of a ‘Patterned’ origin. The rule is 325mm. long. ....................................................... 134 Figure 4.98 The east gable window at Kilgobnet, Inisheer, (L 975 027) has been rebuilt as has the east wall (note numbered stones). ................................................................................................................................................ 134 Figure 4.99 The west doorway of the oratory at Kilgobnet, Inisheer, viewed from the exterior. ............................... 134 Figure 4.100 A view of the same west doorway at Kilgobnet (Figure 4.99) as seen through the east gable window. The vertically emplaced spacing stones in the fabric of the west wall, observed from a distance only, suggest that this wall may have been rebuilt in ‘Romanesque’ times. ........................................................................... 134 Figure 4.101 Temple Benan, Inishmore (L 884 071) viewed from the north-west (the church is orientated with its ‘west’ door towards the north). ................................................................................................................................. 135 Figure 4.102 The ‘west’ (actually facing north) doorway at Temple Benan, Inishmore viewed from the exterior. The stonework is noticeably in Escomb style. ............................................................................................................... 136 Figure 4.103 The ‘west’ doorway of Temple Benan viewed from the inside. .............................................................. 136 Figure 4.104 The modified window in the ‘north’ wall of Temple Benan viewed from inside. The broken external arching stone is just visible. ............................................................................................................................. 136 Figure 4.105 The west gable of St Bhreacáin, Inishmore (L 811 121), shows the roof outline of the early church, a blocked and altered west doorway and, the north-west anta in which the stones are set to the ‘Patterned’ style. ....... 136 Figure 4.106 Interior view (that is, from the east) of the west gable of St Bhreacáin. The change in the masonry fabrics between the early church and its southern extension is very evident. ............................................................ 137 Figure 4.107 Two masonry breaks occur in the north wall of St Bhreacáin viewed here from the north-west. The more pronounced occurring between the extended nave and the chancel. ..................................................................... 137 Figure 4.108 In the extended north wall of St Bhreacháin viewed from the south, Was this ‘aumbry’ once a ‘window’? ...................................................................................................................................................................... 138 Figure 4.109 The visible gable markings of the extended nave above and to the south of the chancel arch at St Bhreacháin. ................................................................................................................................................................... 138 Figure 4.110 The growth of St Bhreacháin, Inishmore. The present building is sketched on the left with dotted outlines of the positions of earlier walls. The outlines are intentionally not to scale and arrows indicate positions of clear wall fabric change. Successive stages in the building growth are also shown. a = west doorway in use, a* = west doorway blocked; b = south doorway; c = early window, c* = early window blocked, to become aumbry; d = anta, when present; e = chancel arch. ....................................................................................................................... 138 Figure 4.111 View towards the south-east interior corner of Temple Chiaráin (L 873 104). The east wall window and its encompassing elaborate string course are likely to be of an early 13th century ‘Romanesque’ date (see text). .................................................................................................................................................................... 139 Figure 4.112 Temple MacDuagh, Inishmore (L 823 104) is viewed here from the north-west to show the western antae and doorway. ........................................................................................................................................ 140 Figure 4.113 The south walls of the chancel (right) and the older nave of Temple MacDuagh display similar style masonry. This is probably because the stones in the lower part of the later chancel include re-used massive stones from the east end of the earlier nave. ................................................................................................................. 140

xx

Figure 4.114 Little still stands of Temple na Neeve, Inishmore (L 824 103). The orientation of the bedding in the lowest four stones in the south-west quoin, nearest the camera, suggests that the church has a ‘Patterned’ origin. ............................................................................................................................................................................. 141 Figure 4.115 The ruins of Temple Soorney, Inishmore (L 866 105) viewed from the south-west. The nearest corner exhibits stones set in ‘Patterned’ style. .............................................................................................................. 142 Figure 4.116 The monastic site of Kilmacduagh (M 405 000) viewed from the nearby road. Beside the round tower is the cathedral with the ruined church of St John the Baptist nearer to the viewpoint. ..................................... 142 Figure 4.117 The north wall of the nave of Kilmacduagh Cathedral clearly shows the fabric change and wall building that has been undertaken above the height of the north-west corbel. .............................................................. 143 Figure 4.118 A view from the north of the west gable of the nave of Kilmacduagh Cathedral. The heights of the gable and the side walls have been raised, probably in ‘Romanesque’ times. Note the gap between the inner and outer gable wall surfaces. .............................................................................................................................................. 144 Figure 4.119 The west doorway of Kilmacduagh Cathedral nave as observed from the west. ..................................... 144 Figure 4.120 The nave west doorway of Kilmacduagh Cathedral viewed from the interior (east). Compare with Figure 4.119. .................................................................................................................................................................. 144 Figure 4.121 The west face of the ruined church at Kiltiernan, Galway (M 437 156). The change in masonry fabric above the level of the top of the lintelled doorway should be noted. ................................................................. 144 Figure 4.122 Inside view (from the east) of the ‘Patterned’ west doorway at Kiltiernan church, with cow for scale. ............................................................................................................................................................................... 145 Figure 4.123 The south-west anta at Kiltiernan ruined church and the double plinth upon which it rests. ................... 146 Figure 4.124 The south-east anta at Kiltiernan viewed from the north (inside) towards the point where the gable wall once joined the south wall (right). Note that the nave south wall is thicker than the anta wall. (Rule 325mm. long). ............................................................................................................................................................................... 146 Figure 4.125 The triangular-headed south nave wall window at Kiltiernan church. This window is believed to be of early ‘Romanesque’ age having been inserted at the time of the chancel extension and the loss of the east window in the nave. ..................................................................................................................................................... 146 Figure 4.126 The west doorway to Banagher church, Derry (C 676 066) seen from the exterior. Note that an architrave surrounds the whole doorway and that the stonework is set in BH style. .................................................. 146 Figure 4.127 The nave plinth, of square section, at the south-east corner of the nave of Banagher church turns beneath the later chancel wall, which possesses a ‘Romanesque’ style plinth. (Rule 325mm. long). ........................... 147 Figure 4.128 Three major variants in wall fabric style are exhibited in this view of the north face of Banagher church viewed from the north-west. They occur principally; low in the nave wall, higher in the nave wall and in the more distant newer chancel. See text for further details............................................................................................ 147 Figure 4.129 A view of the chancel fabric and south chancel ‘Romanesque’ window at Banagher church as seen from the south-west. ...................................................................................................................................................... 148 Figure 4.130 The chapel built by Cormac Mac Cárrthaig (MacCarthy) at Cashel (S 075 409) viewed from the south-east. Clearly of ‘Romanesque’ style, the chapel has many similarities with the church at Bradford on Avon (see Figure 2.68). .......................................................................................................................................................... 150 Figure 4.131 Gallarus oratory (Q 393 048), viewed here from the west, displays no ‘Patterned’ workmanship. ....... 151 Figure 4.132 Temple Cashel oratory (V 369 687) also viewed from the west was many similarities with Gallarus. ....................................................................................................................................................................... 152

xxi

Figure 4.133 Detail of the east window at Temple Cashel. Note that all stones are laid BH and that the window possesses both a protective lintel and a sill. To the right of the pen (135mm. in length) a hole has been drilled through the lintel stone for fastening the hinge of the window. ..................................................................................... 152 Figure 5.1 The inside of the south wall of the older ruined church at Maghera, Co. Down (J 372 342) exhibits an interesting collection of local rock types which include blocks of columnar basalt, boulders of granite and spalls of altered greywacke. .................................................................................................................................................... 154 Figure 5.2 The Burren Carboniferous limestones (Unit 64, Table 4.1), seen here in the walls of Oughtmama East church (M 305 079) as viewed from the south-west, provide distinctly different wall fabric to that seen in Figure 5.1. ................................................................................................................................................................................. 154 Figure 5.3 Walls of partly hammered and shaped blocks of Upper Old Red Sandstone (Unit 54, Table 4.1), seen here in the ruined oratory at Killelton, Kerry (Q 720 100), afford a further wall fabric to those seen in Figures 5.1 and 5.2. .......................................................................................................................................................................... 155 Figure 5.4 Viewed from the west, St Kieran, Glendalough (T 123 967) is built of local phyllite which provides a different wall fabric again to those seen in Figures 5.1 to 5.3. Note that the lower stones in the doorway jambs are orientated vertically. ................................................................................................................................................. 155 Figure 5.5 The south wall of Clonkeen church, Limerick (R 689 547) viewed from the south-west clearly displays masonry of two distinct types (out of view, both terminate in an anta). The two rock types are of different lithologies and the masonry styles are of different ages: that at the lower western end has a number of ‘Patterned’ characteristics; that at the eastern end is possibly 15th century. Which of these two styles Ó’Carragáin (2005a) chose to analyse is uncertain. ............................................................................................................................ 156 Figure 5.6 The inside of the ruined early church at Kilree, Kilkenny (S 497 410) and the west side of the chancel arch reveal the use of at least three stone types and three or more periods of construction. .......................................... 156 Figure 5.7 The north wall of Clonamery church, Kilkenny (S 658 355), 2m. west of the infilled doorway viewed from the north (exterior). This displays worked granite blocks in courses supported by spalls. Compare with Figure 5.8. ..................................................................................................................................................................... 157 Figure 5.8 The wall exactly opposing that shown in Figure 5.7 (that is, the inside of the north wall of Clonamery church 2m. west of the infilled doorway). This fabric is quite unlike that in the external surface, for instance it contains broken boulders and pieces of phyllite, etc. (Rule 325mm. long). .................................................................. 157 Figure 5.9 The south wall of Clonamery church, viewed from the inside (and directly opposite Figure 5.8) is constructed of a further variety of fabric, mainly granite boulders and phyllite, but in good courses. Compare with Figures 5.7 and 5.8. (Rule 325mm. long). ............................................................................................................. 158 Figure 5.10 Lower Old Red Sandstones in a quarry in Mitcheldean, Forest of Dean (SO 671 185), illustrate the manner in which individual bands of rock may vary enormously in thickness. ............................................................ 158 Figure 5.11 Cambrian slates and siltstones in cliffs at Fethard, Wexford (S 801 035) as in Figure 5.10, similarly show changes in thickness in individual beds. .............................................................................................................. 159 Figure 5.12 The inside of the south wall of Killulta church, Limerick (R 434 534) viewed from the north-east (see section 4.5.6). The wall is made up of well-coursed blocks of Carboniferous limestone which have been worked from individual beds. One bed (or course) over the wall length of 4m. (approximately one-third up wall) changes its thickness from 135mm. to 60mm. .............................................................................................................. 159 Figure 5.13 The use of spalls (mainly of igneous rock) to semi-patterned effect in the south, largely granite and basalt, wall of boulders and irregular blocks, low in the chancel of the Maghera old ruined church, Derry (C 855 002). ............................................................................................................................................................................... 160 Figure 5.14 The north wall of the church at St John’s Point, Co. Down (J 528 338) viewed from the east end where the remnants of the north-east anta occur. One megalithic stone placed BVFR (face bedded in the north wall) is visible (rule 325mm.). The dimensions of this stone in this position make it obvious as to why it could not be placed horizontally (BH). .................................................................................................................................... 160

xxii

Figure 6.1 The plan of an early, simple unicameral Irish church with antae is shown. Typically, there is a lack of apparent bonding between the gables and the adjoining walls and often this takes the form as illustrated. ................... 162 Figure 6.2 Kilmalkedar church, Dingle, Kerry (Q 403 062) is viewed here to exhibit its west gable. The antae rise to roof height where they are ‘capped’ with slightly broader stones; they then continue a very short distance in line with the roof. The finials are, however, constructed on top of the gable and are not associated with the antae. There are at least two, and probably three building periods involved in the construction of the west gable wall, the earliest in the wall rising only to the height of the top of the arch jambs. ....................................................... 163 Figure 6.3 Muiredach’s Cross, Monasterboice (O 043 821), at its head carries a ‘church-like’ structure bearing finials. These finials are noticeably on the top of the gables. ......................................................................................... 164 Figure 6.4 Traditionally described as the shrine of St Tighernach, this tomb in the churchyard of Clones church, Monaghan (H 502 256) is constructed with finials. It has been suggested that structures such as this support the contention that they were modelled on existing churches, although Harbison (1999) offered a different view. The tomb possesses no antae and the finials are built on the top of the gable. ..................................................................... 164 Figure 6.5 A map of Ireland to illustrate the position of those churches which are known to possess antae or traces of antae. Those churches detailed in Table 6.1 are illustrated with a solid circle and those not examined are shown with a cross. A few churches for which traces of antae have been claimed, such as Ardpatrick (R 637 208), are omitted because no evidence of an anta could be seen. All church sites are close to navigable waters. Area A is drained by the River Shannon and its tributaries; in Area B the churches may be reached by the river systems of the Suir, Nore and Barrow draining into the Waterford Estuary; and in Area C the River Blackwater flows into the Youghal Estuary....................................................................................................................................... 165 Figure 6.6 The lower portion of the north-west anta at Glendalough Cathedral. Note that the anta wall appears to have been built prior to the gable wall (to the right) and the two walls do not appear to be bonded. ....................... 166 Figure 6.7 The shallow, south-west anta of Kilmalkedar church, Dingle, appears to lack bonding between the anta wall and the west gable wall and like Figure 6.6 suggests a line of weakness in the construction. ...................... 166 Figure 6.8 At Clara church, Kilkenny (S 578 523) the south-east anta wall is slightly displaced from the east gable wall. Bonding between the two walls when viewed from the east appears to be negligible and of the two walls in this view the gable wall seems to be of a more recent date. ............................................................................. 166 Figure 6.9 In the last few years the north-east anta at Sheastown church, Kilkenny (S 544 523) has started to fall away from its adjoining east gable wall. The east end of the site is viewed in this photograph. Figures 6.10 to 6.12 display the relationship in more detail. ................................................................................................................ 167 Figure 6.10 A closer view of the north-east anta of Sheastown church. ....................................................................... 167 Figure 6.11 Viewed from the south the same anta as in Figures 6.9 and 6.10 reveals the limited bonding between the gable and anta walls. Holes created to accommodate two gable wall stones can be observed in the anta wall. ........................................................................................................................................................................ 167 Figure 6.12 The same Sheastown north-east anta is now viewed from the north west and much of the anta wall can be seen to have already fallen leaving it unsupported. ........................................................................................... 168 Figure 6.13 Megalithic face-bedded stones built into the lower portion of the east gable wall at St Mel, Ardagh, Longford (N 204686). Constructed between the antae these stones would probably have assisted in the defence of the church. ................................................................................................................................................................. 169 Figure 6.14 A. A simplified sketch-section in the horizontal plane of a possible arrangement of interlocking stones in the wall behind a north-west anta is shown. Mortar is shown in black and individual interlocking stones are shaded differently. With the exception of the facing stones other stones cannot be removed without complete destruction of the walls. B. To a smaller scale, a similar section is shown of a typical side-alternate quoin. Once the large external stones have been removed others behind them (which would have been difficult to fully interlock) could also be separated from the walls. ................................................................................................................................................ 170

xxiii

Figure 6.15 The north-west anta of Temple Bhreacháin, Inishmore (L 811 121). The anta wall gives the impression of having been built first and being only poorly bonded to the gable wall. The anta is constructed of stones which are ornamentally inserted into the structure in ‘Patterned’ fashion. Note that some of the stones are face-bedded whilst others are edge-bedded (see section 4.5.32 for detail). ................................................................... 171 Figure 6.16 The north-east anta to the early 12th century nave of Dungiven ‘priory’, Derry (C 694 084) has been ornamented with ‘Romanesque’ clasping corners. The rule is 325mm. long. Note that in typical ‘Norman’ style of England, the corner stones have been cut back. ........................................................................................................ 171 Figure 6.17 The west face of Clonkeen church (R 689 547) as viewed from the south-west. Although the lowest part of the west gable wall (and short stretches of the north and south walls) are probably of ‘Patterned’ age, both the west antae have been entirely rebuilt (probably, in part, since the church fell into disuse in the mid-17th century), with some of the stones being relatively modern. The stonework settings of the west doorway are ‘Romanesque’ (possibly partly including re-used earlier stones). ................................................................................. 172 Figure 7.1 Viewed from the north-west, Oughtmama East church, Clare (M 305 079) displays a typical early style plinth. The stones in the quoin above the plinth rise as BH, BH, BVFL, BVFR, but the two highest stones are probably not original. More study of the plinth might confirm that this church was originally of ‘Patterned’ origin. The rule is 325mm. long...................................................................................................................................... 176 Figure 7.2 The round tower at Glendalough (T 123 968) is here viewed from the south-east. ................................... 177 Figure 7.3 The west window at the top of the Glendalough round tower is constructed of blocks of phyllite which are set in ‘Patterned’ style. The jambs, each of three stones, are placed in long and short setting and orientated BVFIA, BH, BVFIA. ................................................................................................................................... 178 Figure 7.4 The south window at the top of the Glendalough tower is similar to the west window (Figure 7.3), but the right jamb appears to have three BH stones between the BVFIA stones. .............................................................. 178 Figure 7.5 A further example of the windows in Glendalough tower, this the east window illuminating the level below the top floor, with single BVFIA jamb stones. The lintel stone attracting the lichen may be of granite. ........... 178 Figure 7.6 The doorway to the Glendalough round tower is largely reconstructed of granite (which attracts lichen). Any lineation in these stones is impossible to determine. ................................................................................. 178 Figure 7.7 Lemanaghan Oratory, Offaly (N 176 270), when visited in 2005 was being rebuilt. Note the pile of stones beside the north wall of the church. .................................................................................................................... 181 Figure 7.8 Pyrford church, Surrey (TQ 040 583) is principally constructed of local ferruginously-cemented gravel, but modern replacement stones in structural situations such as this window consist of starkly contrasting white Upper Jurassic, Portland Stone. ............................................................................................................................ 181

xxiv

ACKNOWLEDGEMENTS This work covers assimilated knowledge and associated studies accumulated over a period of nearly 35 years and the author will no doubt, with regret, omit to acknowledge many who have given him valuable assistance with the research and with his work in the field and elsewhere over this long period. To all of these unnamed persons, I wish to convey my most sincere and grateful thanks. Throughout the British Isles, support in the field, particularly in the provision of help in the location of some of the more remote, ruined and overgrown churches, was generously given by many who remain anonymous. For the more recent studies, in Scotland and Ireland, the author wishes to express his grateful appreciation to the Council of British Archaeology and the British Academy, respectively. Both these bodies provided an element of financial support which helped to alleviate some of the expenditure related to the examination of ecclesiastical sites over wide geographical areas. Certain libraries have over the years provided helpful and cooperative assistance which enabled the author to gain access more readily to many of the more obscure publications cited in this work. In particular, I wish to recall the kind help received from staff of the following libraries: Fleet, Hampshire; University of Reading; Bodleian, University of Oxford; Clare County Library, Ennis, and Angus Records Office, Forfar. The support and patience generously offered by those numerous editors involved in guiding the author’s papers towards publication should not remain forgotten; and at this time, the publishers and staff of British Archaeological Reports must be thanked profusely – they have done much to simplify and ease the progress of this publication. In a number of instances figures from other authors’ work have been included in the present text: in each instance the work is acknowledged. More generally a wide range of maps have been critical in this study. Across the British Isles both Ordnance Survey and Geological Survey maps have been used extensively and the assistance afforded by both British and Irish sources is gratefully recognised. In acknowledging individuals, many will regrettably and unintentionally be omitted for which my sincere apologies. The personally, much admired, late Dr Harold M. Taylor provided both encouragement and friendship during the earliest formative years of this work. Professors J.R.L. Allen, FRS and G. Astill have on occasions sympathetically reviewed early drafts of problematic papers, and discussions and correspondence with Professors Rosemary Cramp and Eric Fernie, Peter Minter, and Dr Warwick Rodwell have always proved valuable. Luke Wormald, Historic Scotland, has kindly researched answers related to Scottish ecclesiastical matters, and correspondence with Conleth Manning, Tadhg O’Keeffe, and Brian Lacey provided some assistance with Irish problems. Never to be forgotten is the support kindly provided by my always patient wife, my children and grandchildren, together with my friends; for it is they that suffered the long intervals of my absence and lack of attention during the protracted periods of researching and compiling this work.

ALWAYS IN MY THOUGHTS

xxv

IMPORTANT NOTE AND AN APOLOGY Much of this study relies on the ability to interpret geological features with which many will be unfamiliar. Building stones throughout the British Isles are most commonly quarried or obtained from rocks which were originally deposited as sediments in water, and in particular, the sea. Such sedimentary rocks, as sands, sandstones, mudstones, limestones, etc; when deposited, are frequently layered, preserving slight variations in their composition and character. This layering, typically described as bedding or stratification, although generally visible in a cliff or quarry on a macroscopic scale, is normally present also on a microscopic scale. It may, therefore, be visible in a block of rock incorporated as a building stone within, for instance, an early church. In the fabric of long-standing buildings, bedding of this nature is often most clearly observed by means of a magnifying glass, for lichen and grime tend to obscure its presence, although at the time of first emplacement in a building it may have been unmistakably evident. In certain rock types other than sediments a similar lineation or planar development may be created by processes such as heat or pressure. These processes will be described as they are encountered in this work. The orientation of rock bedding or lamination within church wall fabrics and structures appears to have been significant to the fraternity of early stonemasons and its interpretation provides evidence of past building styles. Unfortunately, it proves difficult to portray many stone bedding orientations by means of photographs – each stone would have to be seen in its magnified image as originally interpreted by the author. An apology, is, therefore, offered. While figure captions in this work may detail the presence of certain stone orientations they may not always be very clearly apparent.

John F. Potter is a Visiting Research Professor in the Department of Archaeology, University of Reading, Reading RG6 6AB

xxvi

CHAPTER ONE PREAMBLE: WHAT LEVELS OF UNANIMITY EXISTED IN EARLY ECCLESIASTICAL ARCHITECTURE IN THE BRITISH ISLES?

1.1 Humble beginnings In the early 1970s, the author of the present work was invited by representatives of the British Geological Survey to use his field experience to examine the building materials of a Surrey church. Ripley Church (TQ 052 566) had been drawn to the attention of the Survey; with it a request for the identification of one of the rock types that was evident in the fabric of the chancel walls (Figure 1.1). Subsequent examination revealed that this rock was, quite unpredictably, a gravel composed largely of flint fragments. Unexpectedly, under scrutiny, one block of the gravel was observed to include a small fragment of phosphatized fossil ammonite. Adjoining the church site at that time, a temporary working gravel pit had been opened and, on inspection, this gravel, particularly at a depth of 1.5 to 2m from the surface, proved very similar to that in the church walls. Both the gravel in the pit at this depth and that in the church were nearly black, being impregnated with iron oxides and hydrated iron oxides. So began the initial stages of the current work. Clearly these observations raised questions, which in turn required answers. How could a relatively unconsolidated gravel as seen in the gravel pit be used as a building stone? How extensively was gravel used in churches elsewhere in this manner? When and why was such material used for such construction purposes? And, what was the history of the included ecclesiastical ammonite? Figure 1.1 The north side of the chancel of the church of St Mary, Ripley (TQ 052 566), in Surrey. The ferruginously-cemented blocks of gravel included in this rebuilt wall initiated the present author’s studies in ecclesiastical geology. The gravel blocks are mixed with nodular flints and the steel rule is approximately 320mm. in length.

1.2 Early answers 1.2.1 The ecclesiastical ammonite The palaeontological question provided a geologically exhilarating if complex answer. The ammonite (Figure 1.2), an extinct fossil, could be identified as having lived in seas in which rocks of the Jurassic, Ancholme Group, Oxford Clay Formation were deposited. Rocks of this Formation occur in the clay vale of Oxfordshire and Buckinghamshire, and were deposited about 155 million years ago. Some 40 million years subsequent to this date, the area including the Oxford Clay was uplifted to become a land mass and erosive processes carried materials from the land towards the sea, which at the time existed further to the south. That shallow sea, centred about Godalming in Surrey, received and accumulated these materials, in a deposit recognised today as the Bargate Member, of the Lower Cretaceous, Lower Greensand Group, Sandgate Formation (Figure 1.3). The materials deposited in this manner are known to include

derived fossils, including ammonite fragments, from the Jurassic rocks which now occur north of the Chilterns. Our derived ammonite must have rested peacefully in the sandy deposits of the Bargate Member for more than 110 million years before uplift of the land and denudation brought it to the surface again during a period of global warming within our Ice Ages. The next part of the ammonite journey (Figure 1.3), from Godalming to Ripley was to be less distant, but involved travelling from the south to the north of the escarpment of the North Downs. Fortunately, an ancestral and torrential River Wey, very swollen with melt waters from snow and 1

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND ice, must have swept up the ammonite to carry it through a valley in the Chalk approximately where Guildford now stands. Together with worn and broken flints from the Chalk and fragments of cherty sandstone from the Lower Greensand, our fragment of ammonite was to be deposited from the waters of the swollen River Wey before it joined the ancestral River Thames, at a locality close to what was to become the village of Ripley. The gravels, deposited approximately 200,000 years ago, relate closely to the Taplow Terrace Gravels of the Thames Valley. But, one further move was yet to follow, a quarryman seeking suitable stone with which to build a church, in a region where rocks of quality are rare or absent, cut out the block of gravel containing the fragment of ammonite for the mason to use in the chancel wall. In due course the present author was to discover that the stonemason involved was likely to have been an Anglo-Saxon, or a person certainly of that period.

Figure 1.2 A sketch of an ammonite similar to the fragmentary form discovered in the walls of the chancel of Ripley church (x 1.3).

1.2.2 Gravel as a building stone That moderately unconsolidated gravel, could be cut from a quarry face with a sharp implement, and subsequently be used as a building stone, required explanation. The task involves an element of patience for hard pebbles of siliceous, flint and chert within the gravel require to be cut around. (Examples of consolidated gravel, that is, conglomerate, having been ‘cut’ exist in the archaeological record. Specifically, the entrance walls of the Mycenaean, c 1250 BC, vaulted chamber known as the Treasury of Atreus, Mycenae, Greece, provide an example; in this and other instances, the smooth surfaces must have been ground down flat with abrasive stones). Following examination of various working pits in the London Basin, within spreads of flood plain, terrace and fluvioglacial gravels (the last, especially in northern Essex), it became clear that an ‘iron oxide’ pan was frequently developed at the level of the groundwater table (Potter, 1987,167 et seq.). This occurred 1 to 3m. below the surface of the ground and might be up to one metre in thickness (Figure 1.4). The ferruginous layer was normally better developed in the older of the superficial gravel deposits. In one instance (at the now back-filled Sendholme pit, Send, TQ 020 547), the iron oxide took the form of specular haematite, but typically the pan was mainly loosely bound with hydrated iron oxides. As discovered by those that first chose to work this material, it was observed that although friable when moist and in situ, if removed and dried it hardened to provide a remarkably robust ferruginously-cemented stone (Figure 1.5). Typically and largely because they are near to water and yet on well-drained ground removed from the inherent dangers of flooding, areas of terrace gravel deposits provided natural sites for early settlements (Potter, 1997). Early churches were thus similarly sited. In digging the foundations of the church, or burial sites in its vicinity, the early Christians would have encountered iron pan, which with drying would have made suitable material for building blocks. A commitment to the use of such an

Figure 1.3 The route of an ecclesiastical ammonite: depicting the travels of an ammonite from the place of its first preservation as a fossil to its final resting place in the chancel of Ripley church in Surrey. 2

CHAPTER ONE

Figure 1.4 Ferruginous gravel in the wall of a gravel pit, now filled, near Thorpe, Surrey.

Figure 1.5 An example of naturally concreted gravel from Linch Hill (SP 403 040), in Oxfordshire. Bridgland (1994, 7) correlates these gravels with those of the Taplow Terrace in the Middle Thames region. (Photo. courtesy, British Geological Survey, A12429). unorthodox and bizarre rock for building purposes would have tended to be obligatory in the London Basin, for there sand and clay predominate and customary types of building stone are virtually absent. Further afield in England, ferruginously-cemented gravels and sands are used in the earliest buildings wherever there is a similar shortage of suitable traditional building stones. Such a situation tends to occur both in the Hampshire Basin and in East Anglia.

1.2.3 The use of ferruginously-cemented gravel and sand in ecclesiastical buildings The occurrence of iron-enriched gravel in the walls of Ripley church provoked inspection of the neighbouring Pyrford (TQ 040 583) and Wisley (TQ 057 596) churches (Figures 1.6 and 1.7). The material was again present. This in turn instigated an ever-increasing area of investigation. Ultimately, every church in the London Basin, the region between the confines of the Chalk, 3

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 1.6 About 2km. north-west of Ripley church, Pyrford church (TQ 040 583) stands on a marked terrace above the River Wey. The major component of the church wall fabric is ferruginously-cemented gravel. In this instance, observed by a party in the west nave wall. The wall is supported by a later sarsen buttress. the oldest church walls or portions of walls were constructed of ferruginously-cemented gravel, sandy gravel or sand (Figure 1.9; and see Potter, 2001a; 2003a). The evidence revealed that in very many of these early churches an Anglo-Saxon origin had been pre-determined on architectural, archaeological or documentary evidence. In those ecclesiastical structures where rebuilding had confused or obliterated such information the ferruginous material was clearly secondary. It had been re-used at rebuilding and was most frequently then associated with other re-used materials. Furthermore, no recognised PostConquest structure was found to contain the same type of iron enriched material in its original walls (other than in a re-used capacity). At this stage, ecclesiastical geology, to the present author, had ‘come of age’; its value in determining the origins and the dating of churches became obvious. Dissimilar stone types clearly influenced regional architectural styles and were helpfully distinguishable when used in different periods of building. 1.3 Ecclesiastical Geology – a new study evolves The studies described above rapidly evolved into an examination of all the building stones and materials in the walls of the London Basin churches (Figure 1.8) and a wide range of articles resulted (Potter, 1998; 1999; 2000a; 2000b; 2001b; 2002; 2003b; 2004a; 2004b; 2007a; Laming, 2003; Dixon, 2005). Within the London Basin it was recognised that certain rock types were often confined to different periods of architecture. Also, other rock types with particular properties were imported into the region. Possibly polished, a stone might reflect a level of financial opulence either of the church or its patrons

Figure 1.7 The extensively rendered Wisley church (TQ 057 596) is about 3km. due north of Ripley church. In the south wall of the nave removed render reveals ferruginously-cemented gravel adjoining the structure of a later brick window. Chilterns and North Downs, was viewed (Figure 1.8). This widespread inspection revealed that only the very earliest churches in the London Basin, and in particular 4

CHAPTER ONE

Figure 1.8 In the London Basin the present author has examined the external stonework of all churches. The region of the Basin may be defined as the boundary between the Upper Cretaceous (Chalk) and the Palaeogene rocks; this is marked by a solid line. The broken line bounds the area in which all churches have been visited. Major national grid lines are marked.

Figure 1.9 Churches in the London Basin incorporating ferruginously-cemented gravel or sand in their fabric. The lack of frequency in the immediate region of London is believed to reflect the repeated rebuilding of churches in the region related to population growth since Anglo-Saxon times. The broken line indicates the area in which all churches have been examined. (Largely after Potter, 2001a). 5

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND (Potter, 2004a). Notably, the Anglo-Saxons chose to utilize rock materials from sources that were as close to their churches as possible, an observation first recognised by Taylor and Taylor (1965, 12). Both the Romans and the Normans frequently selected stone from more distant sources; the Roman employment of Marquise Oolite (Worssam and Tatton-Brown, 1990) and the extensive Norman use of Caen Stone, both rocks from France, may be cited. The rock lithologies displayed in Anglo-Saxon church wall fabrics are often so close to their source that the churches can be used to delineate approximate geological boundaries (Potter, 2005c). Churches, however, provide by far the greatest corpus of information on early use of stone for building purposes in the United Kingdom. The information revealed proves to be very much more than of restricted geological interest.

1.4 Geology and architectural style 1.4.1 Geological influences and the Anglo-Saxon styles In the present author’s initial studies concerning the churches of the London Basin it became apparent that certain features of architectural style could be related to rock types and, more especially in the earliest churches, the availability of different rocks. In areas of very proximate churches, where the available stone was similar, church features would closely resemble each other. This no doubt must in part have been the result of comparison or the need to better the neighbour, or perhaps reliance on the same mason or ‘architect’. It might apply to a major structure (as a porch, chapel or aisle) or a minor stylistic feature (as the position and/or detail of a window or doorway). It applied to churches of all historic periods and to their subsequent alterations. On a wider geographical scale where rock lithology and availability changed, the architecture, in particular for the Anglo-Saxon churches, would also change perceptibly. Their masons, mainly reliant on local materials, were rock dependent. In an area, for instance, where irregular cobbles (as flints or Bunter quartzites) provided the building stone, standard architectural features of their period such as the pilaster-strip or double-splayed window would virtually disappear. Despite these influences periodic changes in church architecture have long been known (see section 2.1) and the Anglo-Saxon style of architecture has been well recognised throughout England (Taylor and Taylor, 1965).

Outside the London Basin, others have now commenced examining the use of different rock types in early church walls. A detailed scrutiny of the wall stones in the AngloSaxon church of Brixworth (SP 747 713), Northamptonshire has been undertaken (Sutherland and Parsons, 1984; Sutherland, 1990; 2003): in these works not only the diverse variety of rock types have been recorded but also the position of burnt stones. Such studies remain uncommon but further examples exist in the works of Bagshaw (1998) at Leonard Stanley (SO 802 032), in Gloucestershire, and Aldsworth and Harris (1988) at Sompting (TQ 161 056), in West Sussex. In Norfolk, church and other distributions of both Leziate Stone (Allen, et al., 2001) and Carrstone (Allen, 2004) have been examined; Worssam and Tatton-Brown (1993) described the use of the Lower Cretaceous, harder Hythe Beds in Kent; Roberts (1974) outlined the occurrence of Totternhoe Stone in Hertfordshire churches, and in West Sussex, Bone and Bone (2000) described the presence in churches of a further variety of Chalk, known as Lavant Stone. Tatton-Brown (2001) described both the distribution and quarrying of certain Upper Greensand rocks in the Middle Ages. Various helpful texts describe the use of stone in general and well-known in this sphere are works such as Jope (1964), Clifton-Taylor (1972) and Parsons (1990). Allen (2004, 1-3) usefully lists some of these and other works that have contributed to the growth in interest in ecclesiastical geology in recent years. Diana Sutherland (2000, 292-295), provides a valuable definition of the subject of ‘ecclesiastical geology’ and cites further examples of work in the field.

1.4.2 Stone bedding orientations The interesting inter-relationships that were evident in the London Basin between rock types and Anglo-Saxon church architecture influenced the present author to look beyond the Basin, to where rock variation was more dramatic. With regard to one publication (Pearson and Potter, 2002) where the building fabrics of Romney Marsh were examined, the present author was appropriately corrected in connection with the description of one church. Professor David Hinton (2002) properly observed that the ferruginously-cemented, sandy-gravel quoin stones that had been illustrated in Plate III (see Figure 1.10), at the north-east corner of Appledore church (TQ 957 293) were in fact laid in ‘side-alternate’ style rather than as described as by the two authors as ‘longand-short’ style (2002, 101). The error had arisen because the present author, self-trained as an archaeologist, had not at that time, read the original definition of the two terms as specified by Gilbert (1946). The present author in his detailed scrutiny of many thousands of church quoins in the London Basin had noted that on occasions quoin stones were placed with their bedding orientated vertically, frequently this occurred in alternation with horizontally bedded stones (see sections 2.3 and 2.4). This same relationship he had noted was present also in work described as of a ‘long and short’ style, first described by Rickman (1836). The present author had inferred incorrectly that the style name could be applied to all quoins where alternations in bedding orientations

Identification of rock types, particularly where the rocks are in walls and have suffered extensive weathering is not simple (see section 2.2.3). Exact identifications are not always possible. It has to be remembered that the quarry or locality from which a wall rock was extracted, perhaps a millennium ago, is no longer likely to be visible or traceable.

6

CHAPTER ONE

Figure 1.10 The existing north transept to the church of St Peter and St Paul, Appledore (TQ 957 293). The northeast quoin, which in Anglo-Saxon times was probably the north-east quoin of the nave, is composed of ferruginously-cemented sandy-gravel blocks. Many of these blocks are set with their bedding orientated vertically, which the author is demonstrating to a party of people.

Figure 1.11 Detail of the quoin illustrated in Figure 1.10, in which some of the stones exhibit bedding orientated vertically (pen provides scale). ecclesiastical building stones yet to be discovered? A range of papers tackle or answer some of these questions (Potter, 2005a; 2005b; 2005d; 2005f; 2006a; 2006b; 2006c; 2007a; 2007b; 2008b; 2009a; 2008d; 2009b). Chapter 2 of this work attempts to explain and summarize the conclusions of these papers.

were evident, as could be observed in some of the Appledore quoin stones (Figure 1.11), and was unaware that Gilbert’s terminology (1946) was applicable only to stone shape.

The questions, however, continue beyond England. Anglo-Saxon churches had been identified close to the present border with Scotland. These churches were observed to exhibit distinctive Anglo-Saxon geological aspects in their structures. The border with Scotland had been far from static, surely the same styles, in ecclesiastical geology if nothing else, could be observed north of the current Border? This topic is examined both in two papers (Potter, 2006d; 2008c) and Chapter 3 of this work.

1.5 What then of this preamble’s title? The significance of stone bedding orientation in AngloSaxon church building fabric had obviously not been noted previously. This was probably because it tends to require geological experience to correctly interpret the orientation. Gilbert (1946), for instance, viewed only the orientation of the stone block and chose to ignore any relationship that this might have with the bedding. How widespread in England was the particular style of AngloSaxon stone emplacement? In addition to understanding the influence of rock type on Anglo-Saxon architectural styles, it was important to determine the extent of patterned Anglo-Saxon stone bedding orientation. Was it always distinct from the bedding orientations of other archaeological periods? Were there changes over time during the long period of Anglo-Saxon church building activities? Were there other aspects of the geology of

The preamble refers to the possibility of similar architectural styles, influenced undoubtedly by differing geology, extending throughout the British Isles. Chapter 4 addresses this issue as it occurs in Ireland. Whether or not the same ecclesiastical geological influences were effective throughout Christendom during Anglo-Saxon times must be a task for others to explore. 7

CHAPTER TWO MASONRY DETAIL AND STONEWORK IN ENGLISH EARLY ECCLESIASTICAL ARCHITECTURE

each extensively assisted in determining characteristic features of Anglo-Saxon architecture and distinguishing it from English, Norman and later Post-Conquest work. Ultimately, in a most remarkable achievement, Taylor and Taylor (1965) and Taylor (1978) produced a work in three volumes detailing all the then recognised features that were distinctive of Anglo-Saxon architecture and relating these to almost all of the recognised churches of the period in England. Particularly applicable to the present work are Volumes I (pp. 1-15) and III, which detail and then analyse the characteristics of the architecture of the period. The present author will have cause to refer especially to the pages 1-15, for these summarize the state of knowledge at the time and will prevent repeated reference to various earlier works.

2.1 Historical introduction Credit is normally given to Rickman (1817) for the initial attempt to discriminate certain early styles of architecture in the British Isles. His particular reference to early ecclesiastical architecture permitted him to distinguish two church towers (at Barton-on-Humber [TA 035 219] and Clapham, Bedfordshire [TL 035 525]) to which he assigned an Anglo-Saxon origin (Figures 2.1 and 2.2). By 1836, Rickman had extended his list of Anglo-Saxon churches to 20, each in his opinion built prior to AD 1000 (Rickman, 1836, 26). He also detailed certain features that he believed were distinctive of this architectural period: ‘…there is a peculiar sort of quoining…consisting of a long stone set at the corner and a short one laying on it…’ Rickman (1836, 28-29)

2.2 Selection of stone in English early ecclesiastical architecture 2.2.1 Changes in stone choice and influences of transport over time

He was to describe this feature (p. 40) as ‘long and short work’. Other pertinent observations were that AngloSaxon quoins might occasionally involve very large stone blocks and possibly Roman tiles. The churches exhibited a ‘want of’ buttresses and the use of ‘rude carvings … which are clear imitations of Roman work’ (p. 28).

Modern and recent churches are customarily built of materials like brick, concrete, glass and cut and treated timber. With the 20th and 21st centuries, transport by rail, road and ship has enabled stone types used in the fabric of a modern building to have been quarried from virtually anywhere worldwide. In general, only the older churches possess an external fabric principally of stone and, with the exception of certain Roman structures, the older the church the more local the source of the incorporated material.

Very many authors have contributed to the gradually increasing understanding of the stonework and masonry detail that is typical of the earliest periods of English ecclesiastical architecture. The works of Micklethwaite (1896; 1898); Brown (1903; 1925) and Clapham (1930)

Figure 2.1 The tower of St Peter, Barton-on-Humber (TA 035 219) was one of the first examples of ecclesiastical architecture to be attributed by Rickman (1817) to be of pre-Conquest origin. It is here viewed from the south-west. 8

CHAPTER TWO would exceed the cost of the stone itself for distances over 12 miles (19.3 km.). There is a broad consensus, therefore, amongst historians, that water provided the primary means of moving heavy or bulky goods within medieval England (see, for instance, Bolton, 1980; Parsons, 1991). This applied equally to the Anglo-Saxons despite their reliance on local stone. They certainly utilized stone from the shoreline on occasions of the erection of churches near to the coast. This stone was customarily in the form of beach cobbles or boulders (Potter, 1999; 2005d; Pearson and Potter, 2002), and the use frequently involved local movement by water. Where better quality stone was absent locally, the Anglo-Saxons were obliged to seek stone from further afield. For this purpose, small flatbottomed boats along narrow watercourses must have frequently been used (Clifton-Taylor, 1972, 80; Pearson and Potter, 2002, 107). For example, for quality structural work in certain Essex churches the Anglo-Saxons sought durable stone from quarries at Barnack, near Peterborough, and this material reached its ultimate destination (Figure 2.3), in part, by small rivers and streams (Potter, 2005a). For churches in the south of England (Figure 2.3), somewhat larger boats were probably used to convey superior stone from the Quarr Stone quarries, on the north side of the Isle of Wight, at least over the initial part of the journey (see also, Anderson and Quirk, 1964; Tatton-Brown, 1980; Potter, 2006a).

Figure 2.2 Clapham Tower (TL 035 525) was also determined to be of Anglo-Saxon age by Rickman (1817). Again viewed from the south-west, as at Barton the tower is surmounted by a younger, in this instance Norman, belfry.

2.2.2 The consecration of ecclesiastical sites The land upon which ecclesiastical buildings are established is typically consecrated. From a point of view of the building stone this proves to be a matter of special significance compared with the use of natural stone in other buildings. With deterioration of the building or population growth over time, churches have been reconstructed, repaired and often enlarged. Rebuilding most frequently takes place on, or very near to, the sacred land of the early church site (Potter, 1987). To meet the requirements of the clergy many churches from AngloSaxon times have had a chancel added or widened. Church reconstruction nearly always involves the use of additional, and the normally expensive commodity, stone. Any stone of sound quality left from an earlier building, or portion of a building, is destined to be re-used in the refabricated structure. In some instances the reincorporated stone, particularly when its physical appearance contrasts with the new fabric, may be used to give an ornamental effect to a new wall (Figure 2.4). In other cases it may provide courses of stone at specific levels, more especially at the top or base of a newly reconstructed wall or structure.

The varying quality of, and the different means of transport available for, bulk and heavy commodities such as stone, have significantly influenced the distribution of stone types in the past. The Roman network of roads in Britain certainly assisted the builders of that period to extract for their buildings better quality stone from more widespread sources than those normally available to the Anglo-Saxons that followed. Sellwood (1984), for instance, has described the wide variety and distant sources of the rock types used in the walls of Roman Silchester (Calleva). The Romans, furthermore, used ships to transport rocks from France and elsewhere. They presumably had prior experience of use of these rocks. In the walls of Richborough shore fort, Pearson (2003, 47) refers to the Roman re-use from France, of both a Ditrupa Limestone from the Calcaire Grossiere Formation and Marquise Oolite. Similarly, the Normans, and subsequently others, transported by sea, Caen Stone from the Middle Jurassic of Normandy. It is also evident from the rock materials included in church walls that the ease of road transport, or presumably road quality, appears to have again improved in Post-Conquest times. Documentary sources, in particular, make it clear that until the advent of railways, transport overland was significantly more expensive than by water. Of relevance is the widely used statistic (Salzman, 1967, 119) that during the medieval period the cost of transport overland

The extent to which building stone may be re-used was well illustrated in an example cited in Potter (2003a). The church of St Michael, Heckfield (SU 723 605) possesses in its visible walls a profusion of ferruginously-cemented gravel (Figure 2.5), a rock type considered by the present author to have been quarried restrictively for use by the 9

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.3 The Anglo-Saxon use of a) Barnack Stone and b) Quarr Stone in quoins of ecclesiastical buildings in the southern half of England: the details of these churches are presented in Potter (2005b). Figure 2.4 The south wall of the chancel at Greenford (TQ 145 832), in Middlesex, has principally been rebuilt in relatively modern times using cobbles of quarried and knapped flint. Ferruginously-cemented sand and sandygravel blocks from an earlier church, however, have been re-incorporated decoratively in the walls. Anglo-Saxons (section 1.2.3). The church, however, featured no early architectural characteristics and appeared to have been largely rebuilt in the 19th century. A churchwarden’s account book of 1876 was to relieve the author of embarrassment, for it described in some detail the manner in which much earlier (‘very old’) wall material of gravel had been demolished to build new walls. An aspect of church consecration, which relates to the dating of churches, is associated with re-consecration, which may apply at times of partial, often extensive, rebuilding. The Norman Conquest also resulted in many cases of re-dedication. ‘After the Conquest many if not most parish churches were rededicated to saints with whom the Normans were familiar. St Michael was a particularly popular dedication.’ Rowley (1977, 109)

10

CHAPTER TWO

Figure 2.5 Heckfield church (SU 723 605) in north Hampshire is here viewed from the north-west to show the north aisle. Very unfortunately, the south wall referred to in the accompanying text has now been covered in cement render, but the fabric in this north aisle is probably re-used similarly and obtained from the north nave wall. In instances where only the latter re-consecration date is known or recorded, the dating of included fragments or portions of an earlier building may become confused or difficult to date with certainty. The fragments may be distinguishable architecturally or geologically as of an earlier period but placed in that period only with a level of hesitancy, for the historical consecration record is that of a later date.

leaves it impossible to ascertain the stone composition of many church walls. Extensive problems are added to stone recognition as a result of the build-up of soot, grime, lichen and moss on external walls. Internally, access often proves difficult, walls may again be plastered, painted or covered, are frequently dirty or damp, and almost always insufficiently illuminated for clear identification of exposed stonework. Natural light, rather than artificial, tends to be a requirement for precise rock identification.

2.2.3 Identification of ecclesiastical stone types ‘… no description of a church can be considered complete unless it gives the names of the stones of which the church is built, and the sources, local or otherwise, from which the stones are drawn.’ Livett (1893, 149)

Following possibly as long as a millennium or more in a church wall many rock types develop a weathered crust, this may make the rock appear very dissimilar to the rock in its natural fresh state. Weathering patterns require recognition to enable the originally emplaced rocks to be determined. Creating the greatest problem in rock identification is the predicament that, only very rarely, is the source or quarry identifiable from which a particular early building stone has been extracted. Comparison of a building stone with possible precise sources is normally impossible – although a few instances of documented early quarry workings are known (see section 2.2.1). An awareness that rock lithologies may in certain instances vary laterally in as little as a kilometre, and vertically within less than a metre, is also critically important. Furthermore, the same lithology may be repeated at different levels throughout a geological sequence.

This very apposite remark is as appropriate today as it was over a century ago. However, it proves exceptionally difficult to always comply wholly with the statement. The close examination of the stone fabric of buildings is generally fraught with difficulties. Although quality binoculars are of some assistance, without ladders or scaffolding, the identification of stone at higher levels often remains uncertain. The unfortunate, largely Victorian, practice of coating rubblestone church walls with render, pebbledash or plaster (‘one of the least cherished contributions to the pattern of English building’, Clifton-Taylor, 1972; see also Potter, 2000c), 11

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND Geological experience and rock familiarity gained in the field obviously assist in rock identification. Created rock exposures, such as quarries tend to be only temporary (Potter, 1988; 1995). In the present author’s identifications, it has proved vital to peruse past literature in order to ascertain descriptive geological evidence, particularly of durable lithologies no longer exposed. The naming or identification of church building stones is, therefore, difficult. Unfortunately, academic literature contains numerous false claims and omissions related to the identification of rock types in historic buildings. Without wishing to embarrass various distinguished authors, it is possible to cite in recent publications the identification, for example, of cement render as Folkestone Formation sandstone, chert as basalt, or flint as Hertfordshire Puddingstone. Instances exist where a rock which constitutes as much as 30 per cent of a church building fabric have been overlooked. Church stonework is of necessity examined in situ by means of a strong hand lens, for the removal of samples for laboratory analysis is neither practical nor permitted. Before any rock type is named, it should be analytically compared with the same rock from its natural, field occurrence: this may require bringing a geological field sample to the church wall for comparative purposes. The degree of precision with which stone identity and provenance can be pronounced depends greatly on the nature of each material, for some rock types are more distinctive than others. The finer grained rocks are often especially difficult to identify. No doubt, the time will eventually come when occasional small-scale sampling from church walls will be permitted to allow more precise identification of problematic stone types.

Figure 2.6 The tower of St Peter, Ash (SU 897 508) is constructed of sarsen stones: these stones from the detail of their working were used in building the tower about the 15th century (Potter, 1998, 295).

It is always of value to determine the relative quantities of different stone types within church walls for significant variations in quantity may assist in distinguishing walls of unlike ages. Estimation of the relative quantities of different lithologies, even if only judged by eye, can provide a gauge of the comparative importance of each stone type. Those stones more abundant in terms of quantity, especially in early churches, are often the nearest to their source of extraction.

sands, and rocks that have been silicified like sarsens and other silcretes, are exceptional. An example of the manner in which the silcretes were collected and used in the London Basin churches for wall building purposes is portrayed in Figure 2.6. By contouring the distribution of these particular superficial deposits as they occur as cobbles in approximately 300 churches, their occurrence throughout the region is displayed (Figure 2.7). Two varieties of silcrete are involved, those including pebbles of flint (the Hertfordshire and Bradenham Puddingstones), concentrated about the areas ‘P’; and sarsens (‘S’). The sarsens are more widely distributed in number, with three concentrations being present: in the west of the Basin; near Ipswich; and to the north of Guildford, in Surrey. The subject is further elaborated in Potter (1998).

2.2.4 Superficial deposits ‘There is probably no branch of British Geology which has excited more controversy, or has more special literature devoted to it, than that dealing with the superficial deposits’ Salter (1905) In the London Basin, where every church has been examined by the present author, the importance of these unusual rock types as building stones, particularly in early medieval churches, was very evident. Typically deposited approximately within the last 1.65 million years, the vast majority of these rocks remain unconsolidated. Ferruginously-cemented gravels and

To be selected as rubble for early walls the unconsolidated material has to be both durable and of suitable size. Most commonly the size involved can be defined, in accordance with Wentworth (1922), as a 12

CHAPTER TWO

Figure 2.7 Contoured distribution of those churches in the London Basin which exhibit silcretes in their fabric. The distribution has been weighted for quantity of silcretes, as follows: a significant amount (x3); some, more than 10 pieces (x2); less than 10 pieces (x1). Contouring has been based on 10km. diameter centres. ‘P’ indicates the concentration of Hertfordshire Puddingstone silcrete material and ‘S’ those concentrations of sarsens (irrespective of letter size). The map is based on statistics in Potter (1998). ‘cobble’ (of intermediate axis length 64-256 mm.), but ‘boulders’ (>256 mm.) of manually liftable size, if available, can also be utilized. Rarely are loose cobbles or boulders (with the exception of flints) quarried specifically for wall building purposes. If resistant and hard-wearing (as flints derived from the Chalk in the London region) they will weather to the surface. In many instances they may have then been field-picked from the surface of fields for use in walls. Coarse beach gravels often provide a further adequate source of this material, more especially in those early churches and structures built close to the sea (Potter, 2005d: 2005e). In regions where ice action has modified the landscape, glacial erratics or fluvioglacial cobbles carried by rivers, may fulfill the same need for building material (Potter, 2008a). With much of the present work devoted to the more mountainous parts of the British Isles, reference should also be made to the use of cobbles and boulders from the steeper courses of rivers and streams.

the Romney Marsh area (Pearson and Potter, 2002; Potter, 2005d). The derived status of the majority of superficial deposits means that they may have become involved in a certain degree of mixing. For example, cobbles in a mountain stream are likely to have become mixed as a result of the stream or its tributaries crossing different bedrock deposits. Likewise, deposits of glacial origin may be of mixed composition. 2.2.5. Re-use of stone and other materials There is no doubt that the earliest post-Roman builders in Britain relied heavily for their source of stone on any suitable pre-existing buildings. Local Roman structures were salvaged extensively for their stone by the AngloSaxons. Stocker and Everson (1990) described the re-use of stone in Lincolnshire. Roman bricks and tiles (mainly bonding tiles, but inclusive of imbrices and tegulae) were used in the London Basin widely by both Anglo-Saxon and Norman masons. The re-use in the London Basin proves sufficiently widespread for the Roman ceramic materials to be contoured and plotted (Figure 2.8; see Potter [2001b] for a full interpretive explanation). With re-used fragmentary tiles there exists a technical

Anglo-Saxon and comparable early church builders, utilizing rocks close to their churches, in many instances, in preference to quarrying a rock from, for example, a beach cliff, appear to have selected initially, derived cobbles and boulders of the same rock from the beach. This choice is very evident in the churches examined in 13

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.8 A weighted and contoured density map of churches in the London Basin containing Roman tiles (from Potter, 2001b). A weighting of x2 was given for those churches with tiles used in a structural capacity (as quoins) and the distribution was based on 10km. centres. difficulty in the identification of Roman material. Frequently, in church walls, more modern tile materials have been misidentified as of Roman origin. A paper by Minter et al. (2006) discusses this problem and offers assistance in interpretation, whilst a simple technique for distinguishing the textures of bricks and tiles has been described by Potter (2006e).

necessarily by the action of fire; although limestones and certain sandstones may turn red with heat, other rocks, as igneous rocks, are likely to be unaffected (see Sutherland, 1990; for further discussion). Local areas of reddened walling may remain where not too severely damaged, as may be observed at Brixworth (Sutherland, 1990), Stow (SK 882 819) (Taylor and Taylor, 1965, 584-593) and Kenardington (TQ 975 322) (Figures 2.12 and 2.13, and see Potter, 2005d). Severe church fires are of such significance that they may be historically recorded and present an absolute date for the occurrence. Where burnt stones are scattered in their presence they must be reused; their record may provide relative dates, however, for the structure in which they are built (Potter, 2006a, 144).

On occasions re-used stones may have been selected from more than one source and provide, therefore, a variety of rock types. The larger Roman building complexes, as shore forts or cities, may have been erected from a diverse range of rock types, which may have been incorporated into new church construction when they were subsequently excavated from these complexes for rebuilding purposes. Brixworth church in Northamptonshire (SP 747 713), and both Atcham (SJ 541 092) and Wroxeter (SJ 563 082) churches, near to the Roman city of Viroconium, are illustrative of such results (Figures 2.9 to 2.11).

The opportunist collection of beach boulders by early church builders, on occasions, may be revealed by the modifications made to certain boulders by marine organisms. Typically, this involves a boulder displaying encrustations of marine animals (such as barnacles, serpulids or Ostrea attachments) or borings into the surface (such as those of Cliona, Polydora, Hiatella or Pholas). Cobble, or smaller sized material, tends to be devoid of such evidence as it is subject to regular movement by waves. Harder, siliceous rocks will also fail to support these marine life forms. Similarly, modifications of this nature tend to be absent in cobbles

2.2.6 Other building stone variables The stonework of early churches, particularly in church wall fabrics, can often provide important information which is not directly related to either rock lithology or orientation. The presence of burnt stones in a wall may or may not imply their re-use. Nor are all stones reddened 14

CHAPTER TWO

Figure 2.9 Numerous stones from the rocks of the Charnwood Forest area of Leicestershire are included in the lower walls of Brixworth church (SP 747 713), in Northamptonshire. They are believed to have been incorporated from an earlier Roman structure, possibly one similar to the Jewry Wall, in Leicester (Sutherland, 2003, 114). The two dark rocks (left) are of Cambrian, Swithland Slate and the rock top right is an unusual mica rich hornfels, from the Charnian sequence.

Figure 2.10 One of the many Roman stones re-used in Atcham church (SJ 541 092) walls (Potter, 2005f), this example displaying carving.

15

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.11 Wroxeter church (SJ 563 082) possesses walls constructed from a variety of local rock types first used in the Roman site of Viroconium (Potter, 2005f). From the same source is the font which probably previously formed the base to a column (now inverted).

Figure 2.12 Burnt reddened stones from the Ashdown ‘Beds’ sandstone occur around the 13th century tower arch of St Mary, Kenardington (TQ 975 322). Blocks of Hythe Beds and dark ferruginously-cemented sandy-gravel are also present. In 1559, the nave of the church was struck by lightning and destroyed. (See also Figure 2.13 and Potter 2005d for a full description).

Figure 2.13 Kenardington church viewed from the north-east places Figure 2.12 in context. Note that the nave was rebuilt to the south of the axial line of the earlier church.

16

CHAPTER TWO

Figure 2.14 The detail of a septarian nodule from the London Clay displayed in the south nave wall of St Mary the Virgin, Salcott (TL 952 137). The nodule (centre) has been bored by the bivalve Pholas, indicating its collection from a marine site such as a beach (20p coin provides scale). Below the nodule is a flint cobble.

Figure 2.15 Detail of an encrusting marine serpulid on an Ashdown ‘Beds’ calcareous sandstone boulder on the north nave wall of the original Anglo-Saxon church at Lydd (TR 043 209), affording clear evidence of the boulder’s marine origin (Potter, 2005d).

Figure 2.16 The detail of boulders, collected from a marine environment, from the north-east arcade pier of the nave of St Augustine, Canterbury (TR 155 577). Immediately above the lens cap (50mm.) a block of Folkestone Beds sandstone has been encrusted by several serpulids. A grey slab of Hythe Beds calcareous sandstone has been bored by a Pholas, 50mm. to the right of the cap. 17

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.17 This London Clay septarian nodule (below pen) which occurs in the walls of Great Braxted church (TL 851 155) has Roman cement (opus signinum), with included brick fragments, adhering to its surface. A white ‘ring’ of calcium carbonate, representing the point of attachment of an oyster, partly covers both rock and Roman cement. The nodule must have been collected, therefore, from a Roman structure at a time when it was beneath the sea. The Roman shore fort (?Orthona), about 30 km. via the River Blackwater provides such a source. or boulders collected from rivers or streams, where the constant attrition prevents their formation. Examples of marine boulder encrustation and boring are shown in Figures 2.14 to 2.17.

this layering (Figures 2.18 and 2.19). Bedding is poorly displayed in blocks of gravel, but the ferruginouslycemented gravel quoins at Inworth, Essex (TL 879 178) in some instances contained blocks which were clearly set with vertically orientated bedding (Figure 2.20). When the churches in northern Essex were subjected to scrutiny, it became evident that a superior stone had, on the limits of the London Basin, been imported into the region for use in church quoins. The stone was a shelly oolitic limestone, in which the fossil shell fragments were large enough to clearly depict the bedding layers. The rock, identified as Barnack Stone, was unmistakably inserted into a number of quoins with its bedding orientated vertically. Examples of churches which typically display this phenomenon are; Birchhanger (TL 507 228), Strethall (TL 484 398) and Wicken Bonhunt (TL 512 335) (Figures 2.3 and 2.21 to 2.23; see also Potter, 2005a).

Based on the knowledge that the early builders relied on local material, the present author has used early church wall fabrics, as ‘geological exposures’ (e.g. Potter, 2001a), specifically for the purposes of geological research. This use of ecclesiastical stonework, however, falls outside the present study. 2.3 Patterns in early ecclesiastical architectural stonework 2.3.1 Patterns in early quoin stones of the London Basin As intimated in section 1.5, studies of early churches in the London Basin had revealed that a number of churches contained, in certain structures such as quoins, stones that had been emplaced with their bedding or stratification placed vertically. The feature was not common in the region for very many of the earliest churches possessed quoins of the limited types of building stone available, such as, sarsens, and cobbles of flint and calcareous septaria, in which bedding was absent. In a few materials, however, layered sedimentation could be discerned. Calcareous tufa, subsequently hardened to, and properly described, therefore, as travertine (Pentecost, 1993), survived in particular quoins at churches such as Northfleet (TQ 624 741) and Crayford (TQ 512 752) in Kent, and Corringham (TQ 716 742) in Essex, preserved

That each of the churches identified with vertically orientated quoin stones in the London Basin was early in origin there was no doubt. Distinguished authors (as Taylor and Taylor, 1965) had identified undoubted Anglo-Saxon architectural features (e.g. double-splayed windows or long and short quoins) in many, but not all, of the churches. Of the remaining churches, none appeared to have been described as of more recent origin than the early Norman period of architecture. Well documented Norman churches were examined, without success, for quoins which might have included stones with vertically orientated bedding. Initial conclusions suggested that this specific style of stone orientation 18

CHAPTER TWO

Figure 2.19 The 10th to the 12th travertine stones above ground level in the south-west quoin of the tower at Corringham (TQ 716 742) are illustrated. Stone twelve shows the vertical orientation of its bedding (with its face to the right) particularly well. The lens cap is 50mm. in diameter, and the church has been described in more detail by Potter (2005a).

Figure 2.18 The south-west corner of the nave of St Botolph, Northfleet (TQ 624 741) displays stones of travertine set in long and short style with a number of the ‘long’ stones placed with their bedding vertically (Potter, 2006b).

Figure 2.20 Much of the wall fabric of All Saints, Inworth (TL 879 178) is constructed of ferruginously-cemented gravel (Potter, 2005a). The church, here viewed from the north, was first noted for its Anglo-Saxon double-splayed windows by Chancellor (1904-5). Some of the original chancel quoin stones, built prior to the chancel extension, exhibit Anglo-Saxon, vertically orientated bedding. 19

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.21 Part of the north quoin to the west door of the nave of Birchanger church (TL 507 228). Each quoin block of Barnack Stone has been emplaced with its bedding vertical and the bedding planes facing to the right (BVFR).

Figure 2.23 Detail of the lowest three stones in the long and short style, south-west nave quoin at Strethall church of Figure 2.22. From the lowest stone upwards, the bedding orientations (see section 2.4.1) are: BVFR, BH, BVFL. might be confined to the work of Anglo-Saxon masons. To determine the veracity of this observation it was necessary to examine long accepted Anglo-Saxon churches in regions of quality stone. 2.3.2 Patterned stone orientation in Anglo-Saxon churches With a few exceptions where the appropriate stonework is now no longer visible or access proved impossible, each church itemized in the classic work of Taylor and Taylor (1965) has been scrutinized. In nearly every instance, the Anglo-Saxon quoins display examples of vertically orientated bedding in certain stones and the best of these examples are briefly described in Potter (2005b). There are occasions where the stonework bedding is unreadable and requires cleaning to remove thick lichen, but only two clear exceptions to the patterned style were observed and these will be referred to in section 2.10. Some supplementary churches of possible Anglo-Saxon, or even referred to previously as of ‘Norman’ character, were additionally investigated and, where found to exhibit quoins of the distinctive style, they were added to the compilation (Potter, 2005b).

Figure 2.22 The long and short style, south-west quoin to the nave at St Mary the Virgin, Strethall (TL 484 398), in Essex.

In examining superior stonework outside the London Basin it became apparent that the Anglo-Saxon use of 20

CHAPTER TWO

Figure 2.25 The west face of Corbridge church (NY 988 644), Northumberland, here shows the south-west quoin of the original west porch to an Anglo-Saxon church, together with a blocked west doorway. The Roman stones are bedded vertically in the south-west quoin (against which a south aisle has been built) and their orientation has been set to match the orientation in the north-west quoin (also now with an adjoining later north aisle).

Figure 2.24 Part of the south-west quoin of the chapel at Bradwell-on-Sea (TM 031 082) constructed of megalithic stones re-used from the local Roman site. The Barnack Stone used displays the vertically orientated bedding well in the low sunlight. The second stone possesses the hole for a Roman cramp. patterns of vertically bedded stones was not confined to quoins, but actually incorporated, as far as could be determined, into all vertical structural or decorative features of that period. Certainly, the style was invariably unambiguous in pilaster-strips and the jambs of arches and doorways. In window jambs, which in the majority of instances occur high in an Anglo-Saxon wall, the stonework is difficult to read, and confirmation that the style always applies remains uncertain. For readable examples, the vertically bedded style prevails.

In a nationwide study, very rare instances were discovered where the Anglo-Saxon mason chose to sacrifice the customary vertically bedded style (Potter, 2005b, 187). Both in the tower arch jambs at Corbridge (NY 988 644) and the crypt at Hexham (NY 935 641), the ready availability of stone previously used by the Romans, persuaded the Anglo-Saxons to neglect their customary style and to use the Roman stones without rotation. This practice was exceptional, for in the majority of their church structures involving re-used Roman stone, such as Bradwell-on-Sea (TM 031 082), Wroxeter (Potter, 2005f), or even well exhibited in certain Corbridge quoins, the orientation of the bedding was placed vertically (Figures 2.24 and 2.25). More frequently, difficulties arose on occasions where the rock type available to construct a quoin, arch jamb or pilasterstrip, failed to display bedding of any form (see sections 2.3.1 and 2.4.1), and in such circumstances its vertical orientation proved impossible to determine. Typical early churches in Kent exhibiting rubble flint quoins, and therefore, no stone orientations, may be viewed at

The realisation that Anglo-Saxon masons had adopted a style of patterned stone emplacement involving vertically orientated stones, which was visible throughout almost all remnants of their workmanship in England, introduced a significant and new role to ecclesiastical geology. The recognition of various quoins as being of Anglo-Saxon origin allowed the initial dimensions of many structures like naves, chancels and towers to be ascertained for the first time. Norman workmanship, with different stonework, could be much more readily distinguished from Anglo-Saxon work. 21

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.26 The lower part of the flint rubble north-west nave quoin at Coldred church (TR 274 475) in Kent.

Figure 2.28 The top of the flint rubble north-east nave quoin of St Giles, Kingston (TR 198 512) in Kent. Coldred (TR 274 475), Kingsdown (TQ 579 633), Kingston (TR 198 512), and Paddlesworth (TQ 685 621) (Figures 2.26 to 2.28; and see Figure 2.34e). There still remained, however, a requirement for a geological understanding of stone types. The ability to distinguish between different rock lithologies is invaluable and critical in the interpretation of studies involving stone bedding orientation. The presence of stones of unlike lithologies in, for example, a quoin or arch jamb, typically indicates that one of the lithologies is a later replacement. The famous helm tower at Sompting (Figure 2.29), in Sussex (TQ 161 056) affords one of many excellent illustrations of rebuilding of this nature (Potter, 2007b). The recognition that many hundreds of quoins, pilasterstrips and arch jambs would require detailed description involving individual stones, determined the need for a suitable and simple nomenclature for distinguishing the different aspects of stone bedding orientation. Clearly the nomenclature first introduced by Gilbert (1946, 159-162) was not applicable for it referred to stone shape (Figure 2.30), and ignored bedding orientation. The Gilbert notation tends to be confusing in certain aspects, for the almost interchangeable terms ‘face’, ‘end’ and ‘side’ are used, each for two of the surfaces of blocks of ashlar stone. (Note: for a definition of the use of the word ashlar as applied in this volume please refer to the Glossary). He chose to overlook the fact that any rectangular block has

Figure 2.27 The north-east nave quoin of St Edmund, Kingsdown (TQ 579 633). 22

CHAPTER TWO

Figure 2.29 The south face of Sompting church tower (TQ 161 056). On examination, the stone lithologies clearly reveal that above the first stage the quoin and pilaster-strip stones have been replaced with Caen Stone (see Potter, 2007b, for precise particulars).

Figure 2.30 A classification of quoin stones according to their shape, proposed by Gilbert (1946). His term ‘clasping’ referred to a stone which was of no great vertical thickness and square in horizontal cross section (which without removing the stone from the wall is impossible to assess).

six faces or sides. Two of his terms, however, ‘sidealternate’ and ‘face-alternate’, for the structure of certain quoins have been widely adopted (Figure 2.30).

3. ‘…long and short work … where upright pillar-like stones called ‘long’ are contrasted with flat or ‘short’ stones that bond into the wall’. He noted that the majority of authors regarded this type of quoin as a ‘Saxon feature par excellence’, and also that quoins of this nature were indicative of the late Anglo-Saxon period of building, which he defined as period C, or from 970 to 1066 AD. He provided as an example All Saints, Rockland (TL 994 960) (Figure 2.33).

2.4 A new stone orientation nomenclature 2.4.1 New quoin stone nomenclature In 1925, Brown (1925, 24-25, 55) analysed the various treatments in which he believed wall corners had been completed by Anglo-Saxon masons. He described three varieties of wall quoin which be believed provided a conspicuous feature of Anglo-Saxon workmanship. These were with:

Others defining quoins of churches have included Clapham (1930, 108); Fletcher and Jackson (1945); Jackson and Fletcher (1949); Taylor (1961) and Taylor and Taylor (1965, 5-7). The last of these authors illustrated six types of quoin which were believed to occur over the Anglo-Saxon (Figure 2.34) and Norman periods of church building. Unfortunately, actual quoins rarely show the levels of perfection in their construction as demonstrated in the drawings of Taylor and Taylor in Figure 2.34. Figures 2.35 to 2.40 provide six examples of

1. Megalithic stones set alternately north-south and eastwest. He cited Stow church in Lincolnshire (SK 882 819) as an example (Figure 2.31). 2. Random megalithic stones, where the large stones might be re-used Roman material, as at St Mildred, Canterbury (TR 145 575) (Figure 2.32). For these corners he used the expression ‘big stone quoins’.

23

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.31 St Mary, Stow (SK 882 819), in Lincolnshire, is here viewed from the south-west. The south-west quoin of the south transept illustrates the type of quoining of ‘megalithic stones set alternately north-south and east-west’ by Brown (1925), which he recognised as being of Anglo-Saxon craftsmanship.

Figure 2.32 The lowest stones in the south-west nave quoin of St Mildred with St Mary de Castro (TR 145 575), Canterbury. The quoin was described by Brown (1925) as a ‘big stone quoin’ of Anglo-Saxon origin. The stones are of re-used Roman Marquise oolite from France and were described originally by Hussey (1858) as of megalithic character. Potter (2006b) describes the quoins of this church in some detail. The bedding of the lowest two stones is orientated vertically (visible on the second stone).

Figure 2.33 Brown (1925) designated All Saints, Rockland (TL 994 960) as his exemplar of AngloSaxon long and short work, exhibited here in the north-east quoin of the nave.

24

CHAPTER TWO

Figure 2.34 Variants in Anglo-Saxon quoin structure (partly after Taylor and Taylor, 1965, Figure 4): a. Megalithic side-alternate quoins set in rubble walling with each quoin stone equating to several courses of wall stone. This is probably the most common variety of AngloSaxon quoin. Typically, the walling sits upon a plinth of square section with the plinth stones set BH. b. Face-alternate quoins emplaced in much the same style as ‘a.’ above. c. Randomly placed megalithic quoin stones. d. and e. Long and short quoins of two different styles. In d. the stones are cut to shape before they are built into the wall: in e. all the stones are cut back; in most instances this being undertaken after the quoin was built. f. A quoin built of materials like cobbles and Roman tiles. 25

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

2.35

2.38

2.36

2.39

2.37

2.40

Figures 2.35 to 2.40 Typical variations in Anglo-Saxon quoin styles: Figure 2.35 The north-west tower quoin of St Bene’t, Cambrige (TL 449 583), closely follows the long and short style of Figure 2.34e. Figure 2.36 The church of Barnetby-le-Wold (TA 061 090) exhibits in its south-east nave quoin stones set both in sideand face- alternate pattern (Figure 2.34a. and b.). Figure 2.37 The megalithic stones in the north-west nave quoin of Duntisbourne Rous church (SO 985 060) most closely relate to Figure 2.34a. Figure 2.38 Gosbeck church (TM 150 556), in Suffolk, in its north-west nave quoin displays smaller stones largely set in a long and short, but unlike Figure 2.34d. or e., in style. Figure 2.39 The unusual long and short style exhibited in the north-west quoin of Kirkdale church (SE 677 857), in Yorkshire. A tomb slab is inserted in the fabric of the west nave wall. Figure 2.40 The rubble and Roman tile corner to the south-east of the nave of St Peter, Iver (TQ 040 812) possesses no vertically orientated stones (Figure 2.34f.).

26

CHAPTER TWO and therefore, the processes of weathering. Bedding placed in a wall in this natural horizontal manner has been described by the present author, for brevity, as BH (implying Bedding Horizontal). Stones may be orientated to place the bedding vertical, a technique apparently valued by the Anglo-Saxons in quoins for decorative reasons. In this case the face of the bedding (the bedding plane) will be directed to the right (BVFR or Bedding Vertical Face Right), or to the left (BVFL or Bedding Vertical Face Left), as the quoin is viewed (Figure 2.41). This same nomenclature can be applied equally, if required, to Post-Conquest quoins. It is not applicable to materials like bricks or cement blocks for they normally lack any element of internal lamination. In completely correct geological terms, there are strictly six, rather than three possible stone bedding orientations, for inversion of the stone and its bedding, or reversion in the vertical sense, has been discounted in the classification. It is possible to distinguish a stone block placed upside-down (bedding laminae inverted) in some instances by using certain geological features that the rock may possibly preserve. Such evidence occurs in the form of animal burrows, ripple marks, current deposited beds and fossil shell attitudes, as well as other structures. It seems possible, however, to ignore this complexity of inversion for the purposes of architectural interpretation. The craftsmen utilizing the BVFR-BH-BVFL stone emplacement style for work on quoins appears to have used it universally throughout England, applying it to each of the Anglo-Saxon varieties of wall quoin described by Brown (1925) or Taylor and Taylor (1965). Both random megalithic quoins, of large stones frequently of roughly equi-dimensional shape, as at St Mildred, Canterbury, or megalithic side-alternate quoins, display stones with vertically orientated bedding. The ‘long’ stones in the long and short quoins, as might be expected, are always placed with vertically orientated bedding (see sections 2.5.1 and 2.5.2).

Figure 2.41 Using a similar display of quoin stones as shown in Figure 2.30, the stones are annotated with a notation proposed and used by the present author (Potter, 2005b). The notation (BVFR-BH-BVFL) refers to the bedding orientation within each stone and not to the stone shape. typical variation in Anglo-Saxon quoin structure with only the rubble quoin from Iver church (Figure 2.40) failing to show vertically orientated stones. An observation by Jackson and Fletcher (1949, 6-7) is especially pertinent to the present work. They had noted that in placing the ‘long’ stones in an upright position, the stones were not laid on their natural bed, that is, with the stratification approximately horizontal. They attributed this to the ignorance of Anglo-Saxon craftsmen and their lack of understanding of natural, stone weathering processes. This monograph will hopefully show that such assumptions about Anglo-Saxon lack of knowledge are ill-founded.

Certain rock types, as in the south of England, the practically ubiquitous, flint, fail to preserve a bedding orientation because of their mode of formation. In uniform grained or very fine-grained rocks (such as Chalk or Caen Stone) it is often difficult to discern the stratification, for magnification with a high-powered lens is normally insufficient to assist in its identification. On occasions where magnification of this type fails to clearly distinguish the bedding, it is evident that early masons without this facility tended to disregard bedding orientation and stones are set probably without any pattern or specific arrangement.

When a stone is built into a vertical quoin, subject to it possessing bedding or stratification layers, the bedding will be orientated in one of three directions. This statement applies equally to any period of craftsmanship (Figure 2.41). Stonemasons through all church building periods have been aware that, if possible, the orientation of the layering in the rocks selected for use in a wall should be approximately horizontal. In this position, the rocks are typically stronger under the vertical compression present in a wall and generally far less susceptible to water penetration along the bedding planes,

2.4.2 A new nomenclature for pilaster-strip stones The terms ‘pilaster-strip’ (with or without the hyphen), ‘strip-work’, ‘pilaster’ and ‘lesene’ have been used interchangeably in modern descriptive church literature. All refer to a long column of stones, in most AngloSaxon instances of rectangular cross-section (rare instances of flint rubble pilaster-strips are known), which 27

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.43 The south face of Earls Barton church tower (SP 852 638) exhibits an excellent array of pilasterstrips.

It should be noted that the use of the terms face bedded and edge bedded is here used in the same sense as by modern stonemasons. Face bedded stones have their bedding parallel to the face of the wall in which the stones are placed. Edge bedded stones have their bedding parallel to the vertical joints in that wall. Vertical pilaster stones tend to be of BVEB style more frequently than they are BVFB; a phenomenon well displayed in the tower of Earls Barton church (SP 852 638) (Figure 2.43). The reason for the difference in frequency will be discussed in section 2.5.2.

Figure 2.42 Typical stone settings for a pilaster-strip and the notation proposed and used by the present author (Potter, 2005b). This notation (BVFB-BH-BVEB) permits the bedding orientation of the stones to be distinguished.

2.4.3 A new nomenclature for arch jamb stones Vertical bedded stones play an important role in the construction of Anglo-Saxon jambs for doorways, arches and windows. Again, there are three possible stone bedding orientations (Figure 2.44). Horizontally bedded stones (BH) again help to tie the structure into the adjoining walls. The vertically bedded stones in these structures could be emplaced either with the bedding face or plane directed into the arch (Bedding Vertical Face Into the Arch, or BVFIA), or with the bedding laminations parallel to the vertical joints on the inside of the jamb wall (Bedding Vertical Edge Into the Arch, or BVEIA).

often stand proud of the walls of a church. Like the stones in a quoin, the dressed stones in a pilaster-strip can be given a notation related to the orientation of their bedding (Figure 2.42). The principal components of an ashlar pilaster-strip are long stones set with their bedding orientated vertically. These stones may be described as BVFB (Bedding Vertical Face Bedded) or BVEB (Bedded Vertical Edge Bedded). Normally BH (or Bedding Horizontal) stones alternate with BV stones to help to tie the pilaster-strip to the wall, so that a long and short style is developed.

28

CHAPTER TWO There is no doubt that the AngloSaxon craftsman tried to match the stone bedding orientations in arch work with those in the adjoining pilaster-strips and quoins. When viewed from the exterior (or interior) the vertical component of all three structures was emphasized. This visible vertical aspect could be associated with the high window openings, the tall doors and high walls typical of Anglo-Saxon work, together producing a characteristic style (see section 2.7). Within the Anglo-Saxon tradition ‘Escomb fashion’ doorways and arches (Brown, 1925, 54; Taylor and Taylor, 1965, 11) fall into the same pattern, involving only BVFIA and BH stones.

Figure 2.44 Typical stone settings for arch jambs together with the notation proposed and used by the present author (Potter, 2005b). This notation (BVFIA-BH-BVEIA) permits the bedding orientation of the stones to be distinguished. Anglo-Saxon stonemasons appear to have used only the orientations of stone of Bedding Vertical Face Into the Arch (BVFIA) and Bedding Horizontal (BH). Norman masons typically used Bedding Vertical Edge Into the Arch (BVEIA) and (BH).

The evidence suggests that, without exception, the AngloSaxon masons employed only the BVFIA option in their arch jamb structures. This declaration cannot be made dogmatically for many arch structures in Anglo-Saxon walls have proved impossible to scrutinize. The majority of arches are visible in the interior of churches where lighting is poor and plaster coatings prevail. Window arches are customarily too high for close examination in Anglo-Saxon churches and only time and opportunity can confirm the veracity of the statement.

In contrast, Norman (and later) stonemasons used stones in their arch jambs, in which the bedding orientation was set either BH or BVEIA, distinctively unlike the work of the preceding AngloSaxons; a feature which will be discussed further in section 2.13. The Norman tympanum stone reflects this Norman BVEIA orientation. With Roman ashlar buildings reliant on BH orientations for their stonework, the workmanship of the three periods of construction (Roman, Anglo-Saxon and Norman) can be readily distinguished.

2.5 Stone quarrying and extraction 2.5.1 The quarrying of stone That Anglo-Saxon masons had a predilection towards the use of elongated and vertically emplaced blocks of stone, in quoins, pilaster-strips and other features of their churches cannot be in doubt. Use of stones in this manner would appear to result from a desire to emphasize the vertical decorative element in the church façade. That the direction of the bedding or lamination in the elongated blocks should invariably follow their length requires some explanation.

Many of the more important Anglo-Saxon arches are recognised as possessing jambs which are bordered by pilaster-strips, these strips being carried up over the head of the arch as a hoodmoulding, which is typically of a simple square section (Taylor and Taylor, 1965, 11). The pilaster-strips bordering the arch jambs have been found on examination to always follow the BVFIA-BH pattern of the jambs.

The long and short quoin or pilaster is essentially a column. Ideally, a pillar or column requires long blocks of stone placed on end in an upright position. Over most of southern England sedimentary rocks prevail and their variety and nature is such that the units of available rock are insufficiently thick for use in the normal manner (Figure 2.45). The difficulty is illustrated in those 29

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.45 Bedding stratification in sedimentary rocks and the resulting typical shape of hewn rocks (after Potter, 2005b).

Figure 2.46 The church of St Thomas, Winchelsea (TQ 904 174), here viewed from the south-west, was inspired by King Edward I but probably never completed. The nave is absent and only the eastern walls of the transepts remain. In the transept walls, remnants of Viviparus limestone columns are badly weathered (see Potter, 2005d). columns erected in many Post-Conquest churches and cathedrals that are created of polished Viviparus limestones (Purbeck or Wealden ‘marbles’). The ‘marbles’, as a unit of rock, never exceed a thickness of one metre, and inside these buildings, columns of limited thickness can be observed to be set with their bedding vertical. Pillars on the exterior of the incomplete, late 13th century church of St Thomas, Winchelsea (TQ 904 174)

(Figure 2.46), reveal how extensively these limestones decay when set in a vertical orientation and exposed to external weathering (Pearson and Potter, 2002; Potter, 2004a). Anglo-Saxon masons had to be aware of the properties and weathering characteristics of any rock types chosen for features such as long and short work, and it is evident that they selected stone for this purpose with extreme care. Although both Romans (as at 30

CHAPTER TWO

Figure 2.47 Using the ‘grain’ of a rock to extract an elongated, uniform width stone, suitable for a pilaster strip. ‘A’ chisel entry to split stone along bedding plane; ‘B’ second chisel entry to break stone further. Fishbourne, SU 836 048) and Normans employed the ‘marble’, it was apparently rejected by the Anglo-Saxons as unsuitable for building purposes.

masonry purposes have been discussed in some detail by Potter (2006c). The Anglo-Saxon stonemason’s tools of trade for working rocks, whether in the quarry or the church face, appeared to be the hammer and chisel; with the size of these tools being varied according to requirement, as indicated by the variable widths of tool marks on stones.

Anglo-Saxon pilaster-strips are constructed principally of elongated blocks of cut and dressed stone of limited thickness (the proud face). Their distribution throughout the British Isles is interestingly restricted to those areas where the craftsmen of the period could gain access to suitably indurated, well bedded stone: a stone that could be cut or split into long, relatively thin, blocks. Rocks in, for instance, the London Basin which could be hewn into such a shape, are virtually absent, and pilaster-strips are rare. To the north of the River Humber pilasters are no longer present (Taylor and Taylor, 1965, 6): nor again, for different reasons, are there suitable rocks with which to construct these structures. The region of pilaster-strip occurrence relates particularly to that of the Middle Jurassic limestones and similar rocks that can be readily worked. Beyond that region in the British Isles, the rocks, even if sedimentary, tend to be older, harder, more massive, and much more difficult to work into parallelsided blocks.

Anglo-Saxon craftsmen were plainly aware of the ‘grain’ of the rocks on which they chose to work. They appreciated that a rock would split more easily and provide a cleaner, smoother surface if cut parallel to the rock bedding planes (Figure 2.47). In modern quarrying of well bedded rocks such as flagstones, where the rocks are not sawn, a similar practice for the extraction of stone is adopted. Pilaster stones would in preference have been cut to the required thickness in the quarry, and possibly perfected on the ground at the church, before insertion in a church wall. If cut in this manner to provide all stones of equal width, the pilaster could be erected. The ‘long’ stones of the strip then tend to exhibit a BVEB, edge bedded orientation, seen well in the tower at Earls Barton (SP 852 638) (Figure 2.48). In setting the stones in the pilaster, their third dimension would have been hidden from view and they, therefore, required no detailed measurement (Figure 2.47). Each pilaster stone was set slightly proud of the wall face and probably erected slightly prior to the walling stones into which it was eventually to be seen (i.e. the wall stones were built around the pilaster stone as each was set in position).

2.5.2 Working stone for the construction of specialist Anglo-Saxon architectural features The creation of Anglo-Saxon architectural features such as pilaster-strips and quoins involved some shaping of stones and was probably more complex than most would have envisaged. Aspects of the working of stones for 31

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.50 Rock bedding planes and cut backs (after Potter, 2006c). Three ashlar stones of identical shape are shown; their bedding planes (BP) and the traces of these have been illustrated in the three different orientations that are possible. The potential success in trying to create a clean vertical arris and a smooth cut back (by chiselling in the direction of the arrows) will differ in each case: a. Will produce the best result. Having chiselled into the rock to the depth required the cut back should ‘fall away’, parting along its bedding planes. b. Chiselling will tend to weaken the rock and the cut back portion will be difficult to remove. c. A complicated chiselling task and difficult to achieve smooth cut back surfaces.

Figure 2.48 Detail of one of the pilaster-strips in the south face of Earls Barton church (Figure 2.43). Note that both of the ‘long’ stones are placed BVEB.

All rocks are, of course, not well bedded, and in instances where the rock being worked could only be hewn in the quarry into a block approximating to the required shape for a structure like a quoin or pilaster-strip, the rock had to be trimmed subsequently. Rather than careful trimming of the whole face or faces of a stone, with the inherent possibility of breaking the stone, Anglo-Saxon stonemasons devised the technique of creating cut backs. This involved removing unwanted rock only from its visible face. The present author has established certain descriptive terms which may be applied to structures where cut back stones are visible (Potter, 2006c). These terms are illustrated (Figure 2.49), and in this instance the nomenclature has been applied to a pilaster-strip. Again, the masons appeared to use their experience of working rock bedding planes (Figure 2.50), here these possibly being only incipient, so that in pilaster stones with cut backs, very many may be observed to be placed BVFB or face bedded.

Figure 2.49 A typical Anglo-Saxon pilaster-strip in which the stones have been cut back to improve the decorative appearance, annotated with the terms proposed and used by the present author (Potter, 2005b; 2007c) to describe the pilaster. The amount of cut back (on the right of the pilaster) of overlarge stones, has been determined by the proud face width of the narrowest stone (in this instance the lowest stone shown). 32

CHAPTER TWO are extremely rare in Anglo-Saxon churches because of the difficulty of preparing perfectly worked quoin stones with two sides to the correct width over the whole column height. Figure 2.52 ‘a.’ and ‘c.’, further illustrate the cut back difficulties: if the ‘long’ stones are set forward of the building line in the long and short style, the ‘short’ tie stones, in order to match, require to be cut back beyond the width of the proud face of the pilaster ‘long’ stones. Figure 2.53 illustrates the instances where the stones subsequent to arrival from the quarry generally failed to have met the parallel-sided shape and size requirements of the pilaster-strip, arch jamb or quoin. All stones required an element of trimming to achieve an equality of width in their exposed face. In the manner detailed at Wittering church above, the stones would have been cut back to the width of the narrowest stone face to be exposed to view (the proud face). The complexities of creating decorative pilaster-strips and other vertical features probably involved the stonemason responsible for the building, first laying out the stones for the proposed structure horizontally on the ground in order to determine the work required. At Wittering church for instance the two proud faces in certain quoins differ in width and all the stones have been carefully selected and turned to involve the minimum of cutting back and work. 2.6 Cut backs or plaster stops?

Figure 2.51 The Anglo-Saxon quoins in the church at Wittering (TF 056 020) in Northamptonshire are well preserved. The north-east chancel quoin, of long and short style, now abutted by an extension to the north (right) side, exhibits its east face. With the exception of the lowest (BVFR) stone above the plinth, the higher stones are cut back on their south (left) side. The cut back of the higher stones is in keeping with the lowest stone width.

The Anglo-Saxon manner of working stone to create cut backs detailed in section 2.5.2 outlined above, would appear to indicate conclusively that the process was undertaken in order to achieve an appearance of equal width in vertical columns of stone. As early as 1836, however, Rickman (p. 28) had noticed that quoins were sometimes placed slightly in front of the building line of the wall and suggested that this was ‘to allow for the thickness of the plaster’. This view persists today. Brown observing for the first time the presence of the cut back, noted that,

Careful scrutiny of Anglo-Saxon pilaster-strips and quoins exhibiting cut backs, reveals that they were created with the stones generally already set within the wall, and therefore, set forward of the wall face. Having erected a church with corners and pilasters of imperfectly shaped stones, and allowed these to set firmly in their mortar, a plumb-line was used to mark the narrowest stone. Other larger stones in the particular pilaster or quoin were then chiselled back to this line. The quoins of the Anglo-Saxon church at Wittering (TF 056 020) (Figure 2.51) well illustrate the final result (Potter, 2005b; 2009b). It should be noted that in locating the pilaster or quoin (or other structure) stones in front of the wall, the cutting back process was simplified; the depth of the cut back will relate approximately to the distance the stone has been placed in front of the wall face. Furthermore, if only one stone in a pilaster-strip, quoin or arch jamb fails to meet the precise (or nearly precise) requirements of the column width and is set forward of the wall face, all the stones in the column would be set forward to match. Thus quoins of type ‘f.’ (Figure 2.52)

‘…parts of flat slabs that lay along the walls were cut back level with the wall-faces and covered with the plaster so that only that portion of them was visible which corresponded with the width of the uprights.’ Brown (1903, 88) In making this observation Brown was employing the term ‘flat slabs’ to describe the ‘short’ stones of long and short work. He supported the view that plaster was the ‘normal finish to Saxon walling’. Brown illustrated his observation by reference to both Wittering church and a pilaster-strip at Breamore church (SU 153 188) (Figure 2.54), but failed to observe that, in both churches, the ‘long’ stones were also cut back. In 1925 (p. 58), Brown listed a number of localities where plaster of AngloSaxon origin could be observed. Each has been investigated and all cited examples have been found to 33

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.52 Elevations and cross sections of pilaster-strips and quoins, ‘without’ cut backs, in long and short work (partly after Potter, 2006c): a. Elevation of a pilaster-strip in which the ‘long’ stones have been prepared to matching width prior to insertion in the wall. The ‘short’ stones are of variable length (compare with Figure 2.42 in which all stones have been pre-cut). b. Cross section of ‘a.’ along line A-B, with all stones set flush to the wall surface. c. Cross section of ‘a.’ along line A-B, with pilaster face set forward of the wall. Note that in this instance the ‘short’ stones would need to be cut back to match. d. Elevation of a quoin (certain elements of perspective purposefully ignored) in which the ‘long’ stones have been prepared to matching width prior to insertion in the wall. Again the ‘short’ stones are of variable length. e. Cross section of ‘d.’ along line C-D, with all stones set flush to the wall surfaces. f. Cross section of ‘d.’ along line C-D, with the quoin faces set forward of the walls. In this rare instance the ‘short’ stones would have required to have been cut back. 34

CHAPTER TWO

Figure 2.53 Elevations and cross sections of pilaster-strips and quoins, with cut backs, in long and short work (partly after Potter, 2006c): a. Elevation of a pilaster-strip in which the stones required cutting back to match in width. The width of the cut back has been determined by the proud face width of the narrowest stone (the lowest shown). Note the second ‘long’ stone could be described as ‘pear-shaped’. b. Cross section of ‘a.’ along line A-B. Removed cut back stone stippled. c. Elevation of a quoin (certain elements of perspective purposefully ignored) in which the stones required cutting back for stones to match in width. The width of the cut back has been determined by the narrowest points in the column (marked ‘S’). d. Cross section of ‘c.’ along line C-D. Removed cut back stone stippled.

35

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND the cure for this problem and covered the whole wallface…with plaster-like-mortar’. Having ‘made their case’ for the presence of external plaster, Jackson and Fletcher (1949) proposed that pilaster-strips were largely functional, in that they separated areas between quoins into smaller areas which would be less susceptible to weathering. The works of these authors additionally figured and described a number of different styles of pilaster-strips and quoins. On the evidence currently available the case that cut back stones, particularly when associated with quoins and pilaster-strips, provided the Anglo-Saxons with ‘stops for plaster’ can now be refuted. Apart from the arguments presented above, namely that evidence of external plaster in this period appears to have been absent, and that cut backs appear to have been made to create an appearance of regularity of width to vertical ornamentations of stones (section 2.5.2); additional points were made in Potter (2006c).The more salient points were:a. Although there is some evidence that Anglo-Saxon churches were plastered internally, and most churches remain in this state today, pilaster-strips built for internal walls are unknown. Cut back stones on internal walls are extremely rare. b. Only at two churches, Barnack (TF 079 050; tower, second stage) and Stow, do the exteriors of AngloSaxon windows exhibit cut back stones. Had external render been universally applied and ‘plaster stops’ been deemed necessary, the cutting back of window jamb and arch stones and stones in similar structures would have been commonplace. c. Of those churches exhibiting cut backs, only three, Barton-on-Humber (Figure 2.1), Earls Barton (Figures 2.43 and 2.48) and Little Somborne (SU 382 326) (Figure 2.55), carry an external render. In each of these the visible render is relatively modern. Most of the churches carrying pilaster-strips and other vertical ornamentation appear to have never been rendered or plastered.

Figure 2.54 The remnant pilaster-strip on the north nave wall of St Mary, Breamore (SU 153 188). This pilaster was selected by Brown (1903) to be illustrative of the presence of cut backs, although he failed to observe that the ‘long’ stones were cut back. Note that the lower ‘short’ stone has traces only of a possible cut back and that the strip width (proud face) relates to the width of the lowest visible stone which is not cut back (Pen for scale, 140mm.). refer to internal walls or floors (Potter, 2006c). Various sites suggested subsequently by current authors and church historians have also been examined, but the existence of Anglo-Saxon external use of plaster or render, has yet to be revealed (Potter, 2008d). The phrase ‘stop end for plaster’ still remains, nevertheless, in common parlance with reference to pilaster-strips and raised quoins. Examples of external coatings by Norman craftsmen, however, certainly do exist (as Rodwell, 1998).

2.7 Recognition of an Anglo-Saxon ‘Patterned’ style 2.7.1 Introduction The vertical elements of Anglo-Saxon craftsmanship in church construction have now been outlined. They include long recognised features detailed in Taylor and Taylor (1965, 1-15), such as pilaster-strips, long and short work, tall and narrow doorways, high walls, vertical stripwork beside major arches and mid-wall shafts in double belfry windows. Less pronounced are the recognised horizontal aspects of the period; simple plinths, string courses and imposts. To these, elements related to the geology of the stonework can be added; vertically orientated bedding, cut backs and proud face development. Detailed geological examination of AngloSaxon wall fabrics of certain churches of the period indicates that in some instances these too were probably ornamented.

Two separate papers (Fletcher and Jackson, 1945; Jackson and Fletcher, 1949) did much to protract and strengthen the belief in a relationship between cut back presence and plaster. Various fundamental geological errors flaw much of the reasoning in the papers. In the first paper (Fletcher and Jackson, 1945, 18), for example, they argued that many Anglo-Saxon church walls were built of Chalk (for which there is no external evidence [Potter, 2006c]), which made the walls ‘susceptible to penetration by rain and in turn…quickly defaced by frost’. They continued: ‘The Saxons were familiar with 36

CHAPTER TWO

Figure 2.55 All Saints church, Little Somborne (SU 382 326) is here viewed from the east. Details of this church are provided in Potter (2006a). The chancel arch is visible on the east wall, the chancel having been removed. A single pilaster-strip may be observed at the further (west) end of the nave north wall.

Figure 2.56 The inside of the north nave wall of St Peter, Diddlebury (SO 508 853) in which a doublesplayed Anglo-Saxon window is built. The wall exhibits well executed Anglo-Saxon herringbone work, the stones notably lining up with those in the window. Taylor and Taylor (1965, 212-213) provide a full explanation.

37

Figure 2.57 Marton church (SK 840 817), in Lincolnshire, possesses Anglo-Saxon stones with vertically orientated bedding in a quoin in both the lower parts of the tower and nave (Potter, 2005b). The south wall of the lowest portion of the tower, as well as the remaining part of the west end of the nave, is generally thought to display Anglo-Saxon herringbone work (see Taylor and Taylor 1965, 412-414). Stocker and Everson (2006), however, judge all the work to be of Post-Conquest age, whilst accepting (p. 46) that ‘The tower…was attached to a church that was set within an already well-established graveyard’.

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND 2.7.2 Patterns in wall fabric stonework

the Essex coastline, rubble churches of quarried flints may display brown bands of London Clay septarian nodules collected from the beach. In many similar cases, the rock type which provided the colour band occurred in the overlying superficial deposits and the cobbles were field-picked, and the other rock was obtained from the solid geology. Taylor and Taylor (1965, 120 and 631) had noted without further comment examples of such decorative banding. Hart (2003, 86) in his work on East Anglian round towers suggested that certain bands had been created with ‘deliberate decorative intent’. In other instances he thought the banding might have resulted from ‘seasonal building stints’. These views are discussed in Potter (2008d).

First suggestions that the stonework of the walls of Anglo-Saxon churches might have been set in a patterned style were advocated by Waller (1846, 120). He proposed that herringbone masonry was an indication of AngloSaxon work. This proposal was vehemently countered by a range of authors, perhaps most effectively by Brown (1903, 85; 1925, 22). Most today would support the view held by Taylor and Taylor (1965, 13), that herringbone masonry ‘gives no reliable evidence of date’. The herringbone pattern tends to be used whenever elongate, irregular stones of variable thickness are employed in wall building, and it can be seen in work of all periods, including that of modern execution. Anglo-Saxon examples of this type of patterned masonry can be seen at Diddlebury (SO 508 853), in Shropshire (Taylor and Taylor, 1965, 211-214; Potter, 2005f) and elsewhere (Figures 2.56 and 2.57).

The second form of Anglo-Saxon ornamentation to wall fabrics is more difficult to observe and virtually impossible to photograph with clarity. The pattern, created with stones which have been rotated to place the bedding or lineation of the rock vertically, tends to have become hidden under a millennium of grime and lichen growth. To photograph the pattern, the whole wall would require cleaning. In England, just three good instances have been noted by the present author (Potter, 2008d). These combine face bedded (BVFB), edge bedded (BVEB) and normally bedded (BH) stones set in the wall in horizontal patterns (Figure 2.64: the terminology being identical to that used to describe the stone settings in pilaster-strips).

More distinctive of Anglo-Saxon workmanship are two varieties of wall fabric ornamentation recently described by Potter (2008d). These are far from universally present and occur only on occasions where suitable rock variety was available to the Anglo-Saxon mason. Now preserved to various degrees of perfection, horizontal colour bands of stone decorate the walls of a number of churches of rubble construction observed in regions such as East Anglia and the London Basin (Figures 2.58 to 2.63). In all instances at least two contrasting rock lithologies were obtained from sources very close to the church site. Typically, for example, rubble flint walls are ornamented with locally collected, but derived, fluvioglacial or glacial, Triassic Bunter cobbles from the English Midlands (Figure 2.63). Near

The occurrence of this predominantly face bedded pattern at Escomb church (NZ 189 301) (Figure 2.65), in Durham, is of particular interest for much of the stone in the church is believed to be re-used from an earlier Roman site and the present author is unaware of Roman masons ever employing ornamental stonework of this

Figure 2.58 St James the Less, Little Tey (TL 892 237), displays a band of ferruginously-cemented gravel in both its south chancel wall and apse. This band is set in walls of flint cobbles and is disrupted where later windows have been inserted. Potter (2001a) regarded the church as being of late Anglo-Saxon age. 38

CHAPTER TWO

Figure 2.61 The north nave wall of Beeston Anglo-Saxon church (Taylor and Taylor, 1965, 61) also exhibits its originality. The tower band continues along the nave wall as far as the original north-east nave quoin which was also constructed of ferruginously-cemented gravel. Figure 2.59 St Andrew, Heybridge (TL 855 081), like Little Tey, displays a band of ferruginously-cemented gravel, on this occasion high in the first stage of its early tower. A later window again disrupts the band.

Figure 2.62 The church of St Martin, Barholm (TF 090 110) preserves in its south nave wall an Anglo-Saxon blocked doorway in which the stones in the jambs are placed in appropriate vertically bedded style (Potter, 2005b, 209). A pilaster rises above the doorway. To the west, within the porch, is the currently used Norman doorway. Within the Anglo-Saxon work different contrasting Middle Jurassic limestones have been used to provide ornamentation.

Figure 2.60 Ornamentation, provided by ferruginouslycemented gravel, in the round, west, flint rubble tower of St Lawrence, Beeston (TG 328 219). The band continues in the original part of the west nave wall (see also Figure 2.61). 39

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.63 St Botolph church in Hadstock (TL 558 447) has been described in detail by Rodwell (1976), and this south wall of the church nave by Potter (2009b). A band of Bunter cobbles runs across the nave walls, with its base at the level of two Anglo-Saxon double-splayed windows. Note that the band may be traced into the outer splay of the windows.

Figure 2.64 Diagrammatic sketch of an Anglo-Saxon or ‘Patterned’ wall illustrating the use of stones orientated with their bedding vertical. FB = Face bedded, EB = Edge bedded, BH = Bedded horizontally. In the doorway set in the wall, BVFIA = Bedding vertical face into the arch.

40

CHAPTER TWO

Figure 2.65 The south nave wall of Escomb church (NZ 189 301), in Durham, is constructed of blocks of Carboniferous sandstone. These blocks are set in a ‘Patterned’ style of FB-EB-BH stones. Regrettably, largely because of surface grime on the stones, it proves impossible to photograph the various stone bedding orientations.

Figure 2.66 The south wall of the chancel of Jarrow church (NZ 339 652), also exhibits stones emplaced in patterns of FB-EB-BH stones. Despite low sunlight the photograph fails to accurately portray the patterning. nature. Thus, if the stone is re-used, as with stones in the arch jambs and quoins (Potter, 2005b), many stones would have been rotated through ninety degrees before insertion in the church walls.

2.7.3 The ‘Patterned’ style In England, the terminology related to changes in architectural styles is well established. The Anglo-Saxon period of ecclesiastical building (largely described and supplemented in the foregoing pages), is readily distinguished from the Norman or Post-Conquest style, for instance. It can be defined with accuracy to the year 1066, and the Norman Conquest. Church historians, however, believe that changes in style would more probably have been gradual rather than immediate (Taylor and Taylor, 1965, 2; Rowley, 1977, 13), and the length of time to fully instigate the change can be debated (Potter, 2008d: see also section 2.8).

Although, in England, this mainly face bedded style of wall pattern is uncommon and confined apparently to the north of the country (Figure 2.66); where rock lithologies are similar it occurs again in both Scotland and Ireland. It should be noted that the use of face bedded stonework both in walls and other structures provides, in itself, an argument for the absence of plaster or similar cover on Anglo-Saxon churches. The stones orientated in this fashion present a smooth surface to the exterior, one on which the surface cover would typically fail to adhere. 41

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND This volume will describe workmanship of great similarities in early architectural form throughout the British Isles, and yet the Anglo-Saxon domain failed to extend over the whole region. The same styles reproduced in Scotland cannot be Anglo-Saxon; furthermore, there was no Conquest of Scotland. Although, English style building fashions supported by the movement of religious orders may have preceded them, the Normans failed to reach Ireland in number until 1169. The term ‘Anglo-Saxon’ for a similar style of architecture beyond an area once occupied by the AngloSaxons is, therefore, regarded as inappropriate. Terms suitable for the purpose of describing widespread and distinctive building fashions should not be tied to a geographical locality and, until much more extensively studied, not defined by any precise period of time. To meet the requirements, the present author has proposed (Potter, 2008d): ‘Patterned’, to describe those walls or churches, irrespective of size or status, where, in particular, the external faces display the Anglo-Saxon features of England, such as:- stones with bedding or lineation vertically orientated; and where they are present, cut backs (of Anglo-Saxon style), wall banding, long and short quoins, pilasters, etc. ‘Romanesque’, might then be used to describe those walls or churches where the stones in quoins were placed with their bedding or lineation horizontally, the external faces of walls were frequently covered with a render, where the typical and much described, decorative Norman styles of England may be observed, etc.

Figure 2.67 The lower part of the west face of the tower at Cabourne (TA 139 018), in north Lincolnshire. The upper part of the tower was restored and rebuilt in 1872. Although the western corners of the lower component of the tower are constructed with a number of stones placed in vertically bedded orientation (in Anglo-Saxon fashion), the western doorway jambs are built entirely with horizontally bedded (BH) stones. Taylor and Taylor (1965, 127) regarded the doorway, on some of its architectural characteristics, as of late Anglo-Saxon age. The arch includes a tympanum set BVEIA in Norman fashion and Stocker and Evison (2006) placed the whole of this lower portion of the tower in that period (see Potter, 2005b, 198).

In this volume, the two terms will remain in inverted commas, for they will no doubt result in considerable debate. The term ‘Romanesque’, and its derivatives such as ‘Pre-Romanesque’, already possess a multiplicity of meanings (compare Clapham [1930] with any modern edition of Pevsner [as Pevsner and Williamson, 1994, 787]). 2.8 The change in decorative style

last thirty years of the 11th century (Rowley, 1977, 103). Unfortunately, pre-Conquest style ecclesiastical buildings or structures can only infrequently be dated precisely, for documented citation or evidence is rare. By contrast, there is a great volume of accurate historically dated information concerning Post-Conquest buildings.

‘The Norman colonization of England was a protracted process which was essentially aristocratic, ecclesiastical and mercantile in nature …’ Rowley (1977, 11) There has been considerable debate regarding the speed with which the change of architectural style in English churches from the Anglo-Saxon to the Norman style occurred. Taylor and Taylor (1965, 2) wrote, ‘…it is reasonable to believe that some of the surviving buildings in the styles characteristic of the later Anglo-Saxon periods were erected by Anglo-Saxon workmen in the manner to which they were accustomed, but after the Norman Conquest’. The transition period for many authors (as Fernie, 1983, 162 et seq.) has been assessed as gradual. Cathedral and monastic building, however, was certainly proceeding on an unprecedented scale over the

Although it must be accepted that the new Norman ‘Romanesque’ building styles should have arrived, according to geography, later in Scotland and Ireland than in the south of England; within any limited geographical area, the changes in style in the uses of ashlar or subashlar stonework, in particular, appear to have been fairly abrupt. Certain churches in north Lincolnshire, for instance, exhibit style changes that apparently occurred during the period of construction of a portion of the building (Potter, 2005b, 184) (Figure 2.67). 42

CHAPTER TWO Much of the support for a fairly lengthy period of overlap in styles must stem from the inability to distinguish clearly between the different periods of workmanship within particular structures in individual churches. In the past, for example, a quoin built in Anglo-Saxon bedding orientations in its lower part was not necessarily distinguished from later BH laid higher stonework in the same quoin (Potter, 2006c). On architectural evidence alone, the tower of St Nicholas, Cabourne (TA 139 018), partly illustrated in Figure 2.67, for example, has been differently dated by different authors (compare, Taylor and Taylor, 1965, 127, with Stocker and Evison, 2006, 128-129). The detail of the stonework, to the present author, indicates that both sets of authors are partially correct. With evidence for a later insertion of the west doorway absent, it must be inferred that this represents a reflection of the Anglo-Saxon to Norman change in style.

lineation on the rock as is evident in Barnack Stone. The mason working with Caen Stone would normally have been unable to determine its bedding orientation, and it would have been unsuitable for the Anglo-Saxon ‘Patterned’ style of stonework.

Evidence for render or plaster on the exterior walls of Anglo-Saxon churches has yet to be discovered: rather, these craftsmen chose to ornament their church walls with ‘Patterned’ stonework. In England, subsequent to the Conquest, external render was applied to conceal the wall faces and delicate carving of arches and similar structures was to become fashionable.

2.9 Changes in the Anglo-Saxon ‘Patterned’ style over time

Studies of the early stonework of England and Wales by the present author indicate that the change in fashion from the decorated styles of the Anglo-Saxon, to the rendered and differently adorned work of the Normans, was remarkably swift. Fernie (2000, 295) stated, ‘The limestone from Caen was introduced almost immediately after 1066’. There is no evidence to suggest that the new style took much more than 50 years, possibly less, to arrive in Northumbria. Furthermore, it appears that the Scottish Border offered little barrier to the advance of the new style (see Chapter 3).

2.9.1 Introduction Various authors who have attempted to describe AngloSaxon architectural features have, in turn, also tried to determine any significant changes to each of these that may have occurred over the period. The dates of introduction of features have been particularly important matters for speculation. Prominent in considering these matters are the works of Brown (1925) and Taylor (1978). Explicit documentary evidence is unfortunately rare, but Brown divided the Anglo-Saxon period of church building into three: ‘A’, prior to the onset of Viking raids; ‘B’, the unsettled years of the raids (800950); and ‘C’, the following years to 1100. Taylor adopted this division but supplemented it with subdivisions. These divisions were selected as follows (Taylor and Taylor, 1965, 17):-

The new Norman regime wished to impose their influence and they built and rebuilt churches in England in great number. Walls appear to have been constructed both sturdily (and of thicker construction), and rapidly. There are examples of minor errors being made, but these mattered little, for they were to be covered frequently with plaster. The new fashion dictated the disappearance of Escomb style jambs, long and short quoins and similar structures. The Anglo-Saxon styles of vertically orientated lineated stone emplacement, necessitated by these features, were to vanish. Cutting back such stones to shape must have been time consuming. For the stonemason, it would have been simpler and quicker, and involve far less exacting stone collection, to lay the bulk of the stones on their bedding planes. From the masons’ viewpoint the new building fashion must have been popular, and therefore, fairly rapidly adopted. The new techniques involved erecting the building and then permitting those with the necessary skills to concentrate on the intricate carving on clearly visible stonework. Such work would have drawn far more attention and appreciation than had been received as a result of AngloSaxon stone orientation techniques.

Period ‘A’ AD 600-800 ‘A1’ 600-650 ‘A2’ 650-700 Period ‘B’ AD 800-950 ‘B1’ 800-850 ‘B2’ 850-900 Period ‘C’ AD 950-1100 ‘C1’ 950-1000 ‘C2’ 1000-1050

‘A3’ 700-800 ‘B3’ 900-950 ‘C3’ 1050-1100

Taylor and Taylor (1965) attempted to apply a date to each of the churches which they examined using these time divisions, but added the proviso, ‘We wish to emphasize the provisional nature of all the assignments of dates …’. Subsequent authors studying a single or a limited number of churches have in certain instances disagreed with the tentative dates determined by Taylor and Taylor. The vast experience of the Taylors was such, however, that the collective records detailed in their volumes may be used as a basis for comparison. The information that was not available to their studies in the form of particulars related to church stonework can now be applied to their provisional recorded dates.

Stone again may well have played its part in the change of styles. The Normans, familiar with Caen Stone from Normandy, initially chose to import it into England for their building purposes. This Middle Jurassic limestone possesses only incipient bedding traces, its more coherent nature and fine, uniform grain size makes it very suitable for precision carving. In this respect it is superior to its English counterparts, for they tend to be of coarser grain, are frequently oolitic and rich in fragmentary shell debris. The shell debris is apt to impart a marked bedding

43

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND 2.9.2 At what time did the Anglo-Saxon masons introduce stones with vertically orientated bedding into church wall structures? Potter (2005b) provided brief details of 171 churches in which vertically bedded stones could be clearly identified in certain of their quoins. Good examples of similarly orientated stones were identified in a further 16 cases in pilaster-strips and 53 instances in arch jambs. Excluding duplication, 192 individual churches were itemized. If the possible foundation dates of Taylor and Taylor (1965) are applied to these churches, the church numbers are distributed as follows: Period ‘A’ Maximum total 19, Comprised as follows; ‘A’ = 3; probably ‘A’ = 1; possibly ‘A’ = 3; ‘A2’ = 7; ‘A2’ or ‘A3’ = 1; ‘A’ or ‘B’ = 3; ‘A2’ to ‘C2’ = 1. Period ‘B’ Maximum total 9, Comprised as follows; ‘B’ = 4; ‘B1’ = 2; ‘B’ or ‘C’ = 2; earlier than ‘C’ = 1. Period ‘C’ Maximum total 127, Comprised as follows; ‘C’ = 51; probably ‘C’ = 2; possibly ‘C’ = 2; ‘C1’ = 7; possibly ‘C1’ = 1; ‘C1’ or ‘C2’ = 2; ‘C2’or earlier = 1; ‘C2’ = 4; ‘C3’ or earlier = 1; ‘C3’ = 54; possibly ‘C3’ = 2. Taylor and Taylor (1965) preferred not to attempt to date a small number of their churches, this applied in particular to those churches which had been extensively rebuilt or altered. In addition, a further 15 churches were not included in their work Figure 2.68 The church of St Laurence, Bradford on Avon (ST 824 609) presents building work of a variety of periods. The south wall of the chancel illustrated in this figure shows a double-splayed window which from its BH stonework shows no Anglo-Saxon characteristics and has been completely reconstructed. The south-east chancel quoin stones are set in typical Anglo-Saxon style and a number have large cut backs.

A detailed analysis of these figures falls outside the scope of the present work, but the overall numbers in each period approximate in percentage terms to the original distribution of church building dates offered by Taylor and Taylor. Seven, mainly ruined, Kent churches were attributed with an origin in the ‘A1’ category by these authors, but none provides stones in which the bedding orientation can be deciphered. There is a certain amount of documentary evidence for supporting the date allocated to each of the ‘A2’ churches.

recently discussed by the present author elsewhere (Potter, 2006c). Thirty-two examples of churches preserving cut backs were described. With two exceptions, these churches, or their walls, containing the cut backs, were placed by Taylor and Taylor (1965) within the building period ‘C’ or AD 950 to 1100. The first exception was the remarkable church of St Laurence, Bradford on Avon (ST 824 609) (Figure 2.68) which had been dated by these authors as ‘period A2, altered in periods C1 to C3’. Some descriptive details of the church have been given by Potter (2005b, 190; 2006c, 73-74), and some of the cut backs are probably Norman whilst others are more modern. The second exception was the church at Dymock (SO 700 312), which in the present author’s belief provides an example of Norman style cut backs (see section 2.12).

This brief analysis is probably sufficient to confirm that stones were set with their bedding vertically orientated in quoins and arch jambs from the period of the first erection of stone churches by the Anglo-Saxons. Specifically for the 53 arch jambs itemized in Potter (2005b), the building date distribution may be analysed briefly as: Period ‘A’ 9, of which ‘A2’ is 4; Period ‘B’ 4; Period ‘C’ 36; Not known 4. 2.9.3 At what time did Anglo-Saxon masons introduce cut backs in stone?

Walls displaying cut backs occur in churches at Bartonon-Humber, Breamore, Earls Barton and Stow, each dated in the experience of the Taylors to the period ‘C1’. It is suggested (Potter, 2006c) that the earliest use of cut

The answer as to approximately when Anglo-Saxon stonemasons first commenced using stone cut backs was 44

CHAPTER TWO

Figure 2.69 The church at Brixworth (SP 747 713) was erected in a region of England where quality building stone is exceptionally common, and yet, in those structures where Anglo-Saxon masons might have been expected to use such favourable stone (the tower’s south doorway, the south-east nave quoin, etc.) it is absent (see Potter, 2009b). The church is viewed here from the south-west. backs developed about this time, that is, about the beginning of the 10th century. It is considered that their use was probably first employed in the trimming process of the oversize stones in long and short quoins and pilaster-strips. Taylor (1978, 1068) has suggested that these quoins and strips may have first appeared about the second half of the 9th century.

being published, elsewhere. The ‘Patterned’ styles of stone emplacement as described and executed by AngloSaxon masons in England appear to occur wherever the rock sources were suitable. Although the instances of vertical bedded stone emplacement cited in Potter (2005b) do not equate to the total possible number of Anglo-Saxon churches in Taylor and Taylor (1965) it is made clear that the present author has only referred to ‘well-displayed’ examples. As emphasized previously, grime, lichen, insufficient light, height, etc; even tool marking, can make interpretation difficult or impossible.

2.9.4 Can Anglo-Saxon wall patterns be dated? Those Anglo-Saxon polychrome, horizontally banded, wall patterns observed (Potter, 2009b), tend to be present only in, difficult to date, churches built of rubble. All examples were probably constructed over the last 150 years of Anglo-Saxon activity, that is, period ‘C’. In contrast, in England, the small number of churches using face and edge bedded stones in their walls, like Escomb and Jarrow (NZ 339 652), are considered to represent very early Anglo-Saxon construction (period ‘A’). The evidence from elsewhere in the British Isles, however, indicates that this probably reflects only the availability of suitable stone in the north of England, rather than a style restricted to that particular period (see sections 3.3.3, 3.4.2 and 4.7.4).

In England, ‘Patterned’ workmanship, as in rubble walled churches, may be absent. Where it was to be expected, it is absent in just two well known instances. These two ‘Anglo-Saxon’ churches are the famed Brixworth church (Figure 2.69) and the church of All Saints, Wing (SP 880 225) (Figure 2.70). In both instances the Anglo-Saxon origin is not in doubt. The present author has published only brief comment on the former church (Potter, 2009a) and nothing at all on the latter church. In both instances a suspicion remains that the work above ground level has been largely rebuilt at some Post-Conquest date. 2.11 Influences of rock types and lithologies

2.10 Exceptions to the rule 2.11.1 Introduction In this monograph it is not the intention to describe English or Scottish early churches in detail. Much of this information has been published, or is in the process of

All the churches considered in the foregoing sections have been constructed of sedimentary rock, and this type 45

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.70 The apsidal chancel of Wing church (SP 880 225) viewed from the south-east. The apse possesses many unusual features that are atypical of Anglo-Saxon work. The pilaster stones are, for instance, all set BH. of rock is almost universally used in the fabric of English early churches. Variations in the type of sedimentary rock readily available to the mason have in many instances controlled the method and style of wall construction. Churches of rubble, built of young, superficial rock types, such as Bunter cobbles or field-picked flints, are quite unlike those using bedded deposits. Quite probably the round tower (Clifton-Taylor, 1972, 195; Hart, 2003, 13) and even the semi-circular apse, so frequently present in rubble churches, were constructed because of the lack of appropriately consolidated sedimentary rock types.

as an example, their face is somewhat broader than in most churches (Potter, 2006a). This probably relates to the stone, the moderately local Upper Greensand, available for their construction. The fossiliferous, glauconitic, fairly soft sandstone must have been difficult to work into parallel-sided blocks without breakage and, to prevent the rocks breaking, more massive pieces were worked. In the north of England, of the many older and harder rock types accessible, the Anglo-Saxons were fairly restricted and discriminating in their choice for specialist work such as quoins and arches (Potter, 2005b). The geographically widespread Carboniferous Limestone was a rock which was not selected, in this region its massive bedding probably making it unsuitable for these purposes.

Changes in local rock lithology have also influenced the style of building. Anglo-Saxon pilaster-strips are constructed of a limited number of suitable rock types. Of special importance are the well-bedded, durable and relatively easily worked, Middle Jurassic shelly oolitic limestones (e.g. Barnack Stone). In the absence of these limestones, Quarr Stone was frequently used in the southern counties. Discounting pilaster-strips where the rock type, generally because the strip was too high, could not be identified (and a few other instances), strip rock types were identified in just 20 cases. In 40 per cent of these instances the rock employed was Middle Jurassic shelly limestone, and in a further 25 per cent Quarr Stone. Similar and specific rock types were used also in specialist Anglo-Saxon quoins (Potter, 2005b, and see section 2.11.2).

Elsewhere in the British Isles, the choice of rock types differs; the Carboniferous Limestone in Ireland, for instance, includes a wide variety of lithologies (including sandstones). Much more commonly too, both igneous and metamorphic rocks, being locally available, are utilized for building purposes. Created under the influence of pressure, an imposed lineation (normal to that pressure) may result on metamorphic rocks. This planar development, known geologically as rock cleavage, provided the early stonemason working to the ‘Patterned’ style with a phenomenon similar in appearance to sedimentary rock bedding and it could be placed vertically when desired. Igneous rocks, like granite, which cool from a molten state, only rarely visibly display any layered structure (although the craftsman

Even at the detailed level, changes in available rock type have affected a building style. The pilaster-strips (see Figure 2.54) and quoins at Breamore church may be cited 46

CHAPTER TWO Anglo-Saxon craftsman was carefully selective in the use of building stone and the identification of this discerning use enables the extent of church building of the period to be more readily determined. 2.11.2 Analytical summary Table 2.1 attempts to analyse and summarize the broad lithologies and rock types used in the creation of Anglo-Saxon ‘Patterned’ architectural features referred to in this work. A less detailed analysis, which also includes the influences of stonework upon ‘Patterned’ architecture elsewhere in the British Isles, appears in Potter (2008d). The masterful analyses of each of the structures discussed below, provided by Taylor (1978), gives extensive detail of their varied forms but places little emphasis on the influences of rock types. In Potter (2005d, 196-207) those Anglo-Saxon quoins clearly exhibiting vertically orientated stones were itemized (for 171 churches). Many quoins listed were not set in long and short manner. Forty-seven churches were listed as possessing long and short quoins, the quoins of 33 of these churches (70 per cent) were constructed of Middle Jurassic Oolite; in many instances this was a shelly oolite, such as Barnack Stone. The presence of fossil shell fragments assist in imparting a marked Figure 2.71 A distribution map of Anglo-Saxon long and short quoins as bedding to these rocks. Other rock prepared by Taylor (1978, 945) and reproduced by kind permission from types used in the construction of that publication. The figure has been modified by the present author as long and short quoins were: Quarr follows: localities underlined have these quoins created in Middle Jurassic Stone (7 churches; 14 per cent), oolitic limestone; localities with broken underline have the quoins built of Travertine (2 churches, where Quarr Stone. probably the stone was re-used from a Roman source [Potter, 2000a]), and 5 churches where the quoins were of working granite in a modern quarry will be aware of such sandstone, each from a different source (Devonian, layering). Occasionally, limited flow in the cooling Carboniferous, Triassic, and Lower and Upper igneous rock will align crystals as they form to create an Greensands from the Cretaceous), some again being observable lineation. Lineated granites were also used to previously used by the Romans. Figure 2.71, copied by produce the ‘Patterned’ style in Ireland. kind permission from Taylor (1978, 945), illustrates his distribution, of 68 churches which he believed possessed In England, it is probably possible to make a case for the long and short quoins. Anglo-Saxon choice of each rock type which was put to use in their building fabrics. Factors such as abundance, For the vast majority of churches (124 in number) proximity, ease of transport, stone quality, fissility, displaying quoins in which the stones were observed to hardness, etc; all playing their part in the selection. The be placed with vertically orientated bedding, the stones 47

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND Stonework or Architectural Feature Long and short quoins

Regional Occurrence Areas accessible to Middle Jurassic Oolite and Quarr Stone outcrops

Pilaster-strips and pilasters Cut back stonework to produce apparent equality of width

Much as long and short quoins Much as long and short quoins

Vertically bedded stonework in arch jambs (Escomb style) Double-splayed windows

As long and short quoins Typically the south and east of England

Polychrome banding

East Anglia and southeast England

Face bedded banding

Northern England

Megalithic quoins

Northern England, elsewhere typically reused Roman material

Table 2.1

Typical Rock Type or Lithology Well bedded, moderately massive, easily worked limestones and sandstones Much as long and short quoins Moderately massive, easily worked limestones and sandstones As long and short quoins Easily worked to shape rubble, or if siliceous (as flint) wide size/shape variety Rubble church walls, particularly where superficial deposits offer a contrasting colour Very well bedded or lineated rock types as fine grained sandstones, workable as ashlar Hard, massive, difficult to work rock types

Notes See analysis in associated text (Section 2.11) See analyses in Sections 2.5.1 and 2.11 See analyses in associated text (2.11) and Section 2.5.2 See analysis in associated text (Section 2.11) See analysis in associated text (Section 2.11) See analysis in Section 2.7.2 See analysis in Section 2.7.2 See analysis in associated text (Section 2.11)

An assessment of some stonework features which are believed to distinguish churches of ‘Patterned’ style as observed in England. These characteristics are in all instances dependent on the availability of rocks of suitable lithologies

were not placed in long and short style. Rather, their emplacement was side-alternate (see section 2.3.2 and Figure 2.30) in style. In rare instances, as at Crayford church in Kent, a few face-alternate stones might be interspersed within the side-alternate pattern. Subject to the stones possessing a lineation or bedding, and suitable durability, there was no distinction between the popularity of rock types or lithologies used for these sidealternate quoins. It must be recalled that only clearly exhibited examples of vertically orientated bedding disposition appeared in Potter (2005b) to form the basis of this brief analysis and, that there were, therefore, very many other instances where the vertical orientation was difficult to decipher or not well shown. The insertion of stones with their bedding vertically orientated was very obviously Anglo-Saxon practice. Side-alternate stones were presumably much more common than long and short examples because, as has been shown in section 2.5, long stones, although they enhanced the pattern, were difficult to quarry and create.

of size. Brown (1925) used the term megalithic when describing a category of large quoin stones (see section 2.4.1) and it remains in common use today. As applied to the quoins of the 171 churches enumerated in Potter (2005b), in a number of instances, as at St Mildred at Canterbury, and Bradwell-on-Sea, these large stones had been re-used (and reorientated) from earlier Roman sites. On occasions, as at Canterbury and Dover (see Figures 2.32, 2.72 and 2.73), the stones were originally imported from France. Megalithic stonework always involves rock types that are typically massively bedded and, more especially, difficult to work. Such rocks in England tend to occur in the geologically older deposits like coarse Carboniferous sandstones. Certain sandstones in the Hythe Beds of Kent and Sussex also break into megalithic blocks. The 28 churches which Taylor (1978, 920) showed exhibiting stone pilaster-strips are only a few more than those described in Potter (2005b, 208-209). The smaller group, of 20 strips cited earlier in this Section, only includes those in which vertically bedded stones are clearly evident. The distribution by stone type in these 20 churches is very like that for long and short quoins.

Many stones used in ‘Patterned’ style workmanship have been described as ‘megalithic’. Unfortunately, the term, meaning ‘large’ or ‘very large’ has no precise definition 48

CHAPTER TWO

Figure 2.72 The church of St Mary de Castro, Dover (TR 326 417) is here viewed from the south to show the adjacent Pharos and Dover Castle (left). The existence of cut back stonework in ‘Patterned’ churches has been described by Potter (2006c). It was observed by the author in just 29 Anglo-Saxon churches. The importance of Middle Jurassic Oolite is again revealed, in 19 churches (66 per cent) the cut backs occur in this type of stone, and in only one of these instances (Stow church) is the rock not very rich in the fossil shell fragments which assist in creating the rock bedding planes. Quarr Stone (4 churches) is again the next most abundant rock to be involved. Moving down Table 2.1, Escomb style jambs, described in Potter (2005b, 209-213), continue with much the same rock use pattern. These jambs are listed in 33 church instances and Middle Jurassic Oolite is incorporated into 16 of these. On this occasion a variety of Carboniferous sandstones make up the next largest group of rocks employed (9 instances, four being of re-used Roman material). Quarr Stone Escomb jambs occur in three churches. At Dover, St Mary-in-Castro church (TR 326 417) (Figure 2.72), the white jamb stones of this type in the south nave doorway are of the re-used Roman, imported stone, Marquise Oolite. The rock types used in creating double-splayed windows are very dissimilar to those used in the stonework features listed above. Taylor (1978, 841), depicted 78 churches which possess windows of this type, with kind permission, displayed here as Figure 2.74. Fifteen of these are described as being constructed of dressed stone. An accurate analysis of the rock composition of doublesplayed windows would prove complicated, for the majority of these windows are set too high for easy access and scrutiny. At least half of those viewed are built

Figure 2.73 Detail of the north-west quoin of the nave of St Mary de Castro, Dover. Although this part of the quoin is built in ‘Patterned’ style, the stone is of a re-used Roman origin. The large white stone is Marquise Oolite from northern France. See Potter (2006b) for a full explanation. 49

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 2.74 This distribution map showing the concentration of Anglo-Saxon double-splayed windows in south and south-east England was prepared by Taylor (1978, 841). Localities underlined involve an element of dressed stone in the window(s) concerned. Permission to reproduce this figure is gratefully acknowledged. of flint rubble (flint being a fairly distinctive rock type to identify), often this is associated with fragments of Roman tile and field-picked cobbles of other material. The cobble or rubble size tends to be variable in the window splays in order to have permitted ease of construction. In those double-splayed windows built of dressed stone (and in some instances only the arch is of worked stone) on occasions the stone is re-used.

2.11.3 Rock types in other Anglo-Saxon ecclesiastical features In his work of 1978, Taylor made analyses of the varieties and numbers of a significant range of features which he believed were distinctive to Anglo-Saxon church architecture. These included aspects like belfry openings and their mid-wall shafts, single-splayed windows, string courses, and hoodmouldings; all elements of style which were no doubt influenced in their 50

CHAPTER TWO

Figure 2.75 The distribution map of hoodmouldings displayed in Anglo-Saxon churches as compiled by Taylor (1978, 931), from which the map has been reproduced. In the south-east of England these features are absent. The majority of the localities relate to the proximity of the Middle Jurassic limestones. presence or structure by the variability of their local building stones. Unfortunately, the height of these features made any examination of stonework, involving detail such as rock type or bedding orientation, unrealistic. Distributions, where illustrated by Taylor (1978), were certainly regional, as is illustrated in his distribution of churches with hoodmouldings, here shown as Figure 2.75. It is noticeable that this variety of ornamentation is completely absent from south-east England where the local stone types available would have been unsuitable for their construction.

The plinths to churches and their styles were also enumerated by Taylor (1978, 964-966). Except in those visible instances examined by the present author where they were of rubble structure, plinth stones were apparently laid BH (although far from every visible plinth stone has been examined). This would seem to indicate that the Anglo-Saxon masons were very aware of this orientation as advantageous in the prevention of water penetration. At the foot of early walls, where plinth stones are generally buried, but on occasions partially or wholly exposed, rock identification is difficult, for the 51

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND namely Northumbria, Mercia and Wessex (with Kent and East Anglia), has led various authors to propose that the styles of church building in each area was traditionally different (Fernie, 1983). Peers (1901) believed that a number of Kent churches (actually four in Kent, together with South Elmham [TM 309 826] and Bradwell [Figure 2.24]) showed strong resemblances to each other, and Clapham (1930) identified groupings of like churches in Northumbria and south-east England. Other authors, such as Taylor (1978) – who made the most extensive of surveys (Taylor and Taylor, 1965) failed to distinguish these differences across England – but tended to relate stylistic differences in Anglo-Saxon churches as a whole to stone availability.

Figure 2.76 This distribution map of the presence of churches of Anglo-Saxon origin with eastern apses in England was also compiled by Taylor (1978, 1029). Only the small east church at Hexham lies north of the Wash. Taylor (1978, 1028) makes the point that the apsed churches extend ‘widely beyond the early kingdom of Kent’. The evidence is strongly in favour of their origin being related to the influence of the regional geology, and to areas where stone in the form of rubble predominates. stones are more than normally covered in earth and grime. There are certain aspects of Anglo-Saxon church architecture examined by Taylor (1978) where any relationship with the local stone type seems improbable. This must surely apply to features like church size and plans. However, even church size probably relates to the agricultural productivity of the locality, influencing the population, its size and financial stability. This, in turn, relates via the soils, to the geology.

Although Fernie (1983) devoted a chapter to each of the kingdoms, jointly the chapter observations were used to introduce the currently visible forms of various early Anglo-Saxon churches which had been built over the two and a half centuries between the arrivals of St Augustine and the Danes. These range in form from the larger basilicas, via those churches possessing a crypt, or a single porticus (or more), to those with an eastern apse. In his summary Fernie (1983, 7273), concluded that most of the styles were, from the evidence available, held in common; with only the Kentish group retaining its individuality. He continued: ‘This is chiefly due to the combination of the apse and the triple arcade, features which, although not as widely attested as previously thought, nonetheless represent the commonest forms in the kingdom.’ Fernie (1983)

A third ‘characteristic’ of the early Kentish churches, ‘the distinctive use of Roman masonry’, Fernie (p. 73) dismisses with the further observation, ‘(this) may be due to nothing more than what was available for re-use.’ Figure 2.8 confirms the veracity of this remark. Churches possessing apses are apt to occur wherever the wall building fabric is rubble (see section 2.11). There is no doubt that such churches are particularly abundant in East Anglia, especially in undated, small Anglo-Saxon village churches (Figure 2.76). They are present in a number of early Kentish churches where quality building material was not available. Certain churches, like those in

2.12 What of the Anglo-Saxon Kingdoms? Awareness that England prior to the onset of Danish invasions was a mix of many small kingdoms which by the 7th/8th centuries had consolidated into just three, 52

CHAPTER TWO Canterbury, of St Martin and St Paul (TR 158 575), referred to by Bede (Figure 2.77), as well as St Peter and St Paul (TR 155 577), are often described as possessing an apse for which no structural evidence has yet been found (Taylor, 1961). Possibly the most significant feature of eastern apses is that in many instances they tend to have been replaced, generally in late Anglo-Saxon times, by a rectangular chancel. Just two, possibly three, of the early Anglo-Saxon churches in Kent exhibit traces of a triple arcade.

[‘Purbeck marble’], particularly as a polished stone, as for fonts), and refrained from using rocks like ferruginously-cemented gravel. They continued building churches with materials such as flint in regions like East Anglia where alternative stone was scarce, but chose to build extensively elsewhere with ashlar stone laid in courses. This move towards the use of squared stone was, in some instances, remarkably demanding. The modification of sarsens to moderately rectangular blocks must have been difficult, but for example, the late Norman, squaring of large septarian nodule boulders from beaches adjoining the London Clay, for the curtain walling of Framlingham Castle (TM 286 637), in Suffolk, was an extraordinary achievement. In some circumstances stone, where sufficiently soft, was sawn. Brick manufacture recommenced in the Essex region and bricks appear in certain churches (e.g. Rodwell, 1998).

When the building fabric of Anglo-Saxon churches is reviewed, it is clear that the supply of different stone types, including the availability of stone which could be re-used, determined most of the characteristic features of Anglo-Saxon churches such as double-splayed windows, pilaster-strips, etc., (See section 2.11.2 and Table 2.1). Different stone availability had far less control on major structural aspects like church size, plan, shape and constituent compartments (chancel, aisles, porticus presence, etc.). Variation in these major influences on church shape were no doubt related to factors like the level of financial support from the church patron(s) and the tendency to model a church, as for instance the basilica form, on earlier known buildings (in particular those of mainland Europe and the Middle-East). It is clear that in these respects Taylor (1961; 1969), recognised that the earliest Kentish Anglo-Saxon churches had some special significance.

Whereas the Anglo-Saxon craftsman appears to have been reliant on using local stone of appropriate quality; the Norman and subsequent builders, using dressed and sometimes polished stones, were more influenced by the affluence of the patronage of the church. Changes between those rock types employed for ecclesiastical architecture pre- and Post-Conquest are numerous, and somewhat variable according to stone availability, in different regions. 2.13.3 Stone orientation and cut backs in Norman churches

2.13 The Norman ‘Romanesque’ masonry styles Typically Norman masons in England resorted to placing the vast majority of the stones which they used in the orthodox horizontally bedded (BH) style. Where stones were to be carved, they frequently, however, placed the stone so that the easiest face to carve, the bedding plane, was used for this purpose. This situation arose most frequently around arches and seems to universally occur in the Norman tympanum, which in terms of the arch is placed BVEIA (Bedded Vertical Edge Into the Arch). The stonemasons were by this means provided with a smooth, equally-resistant surface to carve. Where arch jamb stones are carved some of these may also be placed BVEIA. Quoin stones are not customarily carved and are, therefore, laid horizontally.

2.13.1 Introduction Both the architectural features (e.g. Browne, 1907; Clapham, 1934; Rowley, 1977; Fernie, 2000;) and their decoration (e.g. Clapham, 1934; Zarnecki, 1951; 1953; Zarnecki, Holt and Holland, 1984) in Norman churches in England (and in Ireland, Henry, 1970; O’Keeffe, 2003) have been extensively described (see also ‘The Corpus of Romanesque Architecture in Britain and Ireland’; http://www.crsbi.ac.uk). This period of building is possibly the most readily recognised by the layman. To date, in depth ecclesiastical geology of this and subsequent periods of architecture have been virtually ignored. What follows in this section touches on certain observations made by the present author, throughout the British Isles, in attempting to distinguish the AngloSaxon ‘Patterned’ styles from those of the Norman ‘Romanesque’. Without detailed, prolonged and thorough study of all Norman workmanship many of the observations briefly made here must remain tentative.

In coursed ashlar work the Norman mason was not averse to inserting the occasional vertically bedded stone. These stones were normally placed in edge bedded style (Figure 2.78). In using stones of similar size in courses there was a tendency for the ‘perpens’ (the vertical mortar joints) to coincide occasionally, vertically between courses; an occurrence which would have weakened the wall. To avoid this happening, a stone would have been broken to an appropriate width and inserted in the wall. It would have been most readily broken along the bedding plane and inserted, therefore, in edge bedded fashion. In the present author’s view, this use of edge bedded stones provides a critical means of determining the age of construction of round structures. As the Norman mason, laid a course of stones in a circular structure the first stone laid would be approached by the last and the final

2.13.2 Examples of the use of stone References to the Norman extensive use of imported Caen Stone in England, and the reasons for this use, have already been made (see section 2.8). The Normans introduced the widespread use of some rock types (for example, Upper Greensand from the northern Weald, [Tatton-Brown, 2001]), re-introduced others (for example, the late Norman use of Viviparus limestone 53

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND gap was normally filled by an edge bedded stone. Stair turrets and round towers tend to exhibit this incidence particularly, for as they taper upwards they tend to involve a different and decreasing number of building stones in each erected course. Anglo-Saxons, building with more irregularly shaped stone blocks and with variable thickness of mortar, tend to use this technique far less regularly.

The clasping quoin and broad shallow pilaster are features found in certain Norman churches. Where these are created in coursed ashlar stone they are apt to result in the Norman ‘Romanesque’ variety of cut back (Figure 2.78). Where the horizontally bedded, bonded ashlar stones in the proud faces of such structures are too large they are cut back. Examples of workmanship of this type may be observed in the broad pilasters (Figure 2.79) and in the south-east clasping quoin of St Mary, Dymock (SO 700 312). In the south-west quoin of St Nicholas, Worth (TQ 302 362) it is possible to compare Anglo-Saxon stone insertion with that, higher, of Norman construction (Figure 2.80). The Anglo-Saxon church of Bradford on Avon exhibits in much of its decorative workmanship Post-Conquest style cut backs (Potter, 2006c).

Norman wall construction generally differed from that of the preceding Anglo-Saxon, ‘Patterned’ period. Apparently erring on the side of over provision their walls are thicker. Whereas the ‘Patterned’ wall is prepared with quality mortar and its stones are wellbonded throughout the wall’s thickness, the later ‘Romanesque’ wall was frequently erected with a rubble infill which contained limited mortar. This is especially the case as the Norman masons improved their skills in the hewing and cutting of ashlar stone, so that the ashlar material provided the ‘strong’ skin to the rubble core. Fernie (2000, 295) provides examples of failures in this type of construction.

In concluding these brief comments on the English Norman styles of stonework it must be emphasized that only a very small proportion of the existing available architecture of this period has been examined. A much more extensive examination of material of this age would be necessary to establish firm rules regarding the use of stone by these Post-Conquest masons.

Figure 2.77 Much of the south wall of the chancel of the early church of St Martin and St Paul, Canterbury (TR 158 575) is displayed in this figure. The doorway with the heavy lintel of very glauconite-rich Hythe Beds sandstone was almost certainly originally quarried by the Romans from close to Hythe. It is believed to have served as an entrance to a south porticus. The round headed doorway constructed of Roman tiles is regarded as a later Anglo-Saxon insertion. The south-east nave quoin (left) is also constructed of re-used Roman material, the white stone blocks being of Ditrupa limestone (Calcaire Grossiere) from France.

54

CHAPTER TWO

Figure 2.78 A typical Norman ‘Romanesque’, bonded ashlar wall with a broad shallow pilaster and a clasping quoin. Where they might have overlapped the proud faces, stones will have been, of necessity, cut back. The vast majority of the stones in the wall and the structures will have been placed with their bedding orientation horizontal, but occasionally where a shorter stone is required it may be cut and have an edge bedded setting (after Potter, 2006c).

Figure 2.79 The south wall of the nave of Dymock church (SO 700 312), showing a Norman pilaster in sandstone of Lower Devonian age (Lower Old Red Sandstone). All the stones are set BH and where they would have extended beyond the pilaster width they have been cut back.

Figure 2.80 The south wall of the nave at St Nicholas, Worth (TQ 302 362). In the upper portion of the southwest quoin, the proud face of the quoin suddenly changes in width: at this point the quoin and adjoining wall change from Anglo-Saxon to Post-Conquest in fashion and the cut back stones are of different style (see Potter, 2006c for detail). 55

CHAPTER THREE EVIDENCE OF ‘PATTERNED’ MASONRY STYLES IN SCOTLAND AND THE ISLE OF MAN

additional stonework features, related especially to stone orientation, were distinctive of Anglo-Saxon or ‘Patterned’ workmanship. Similar orientations had been recorded in a limited number of early churches from Wales, where the rock types more closely resembled those in southern Scotland.

3.1 A Scottish Enigma The existence of early Christian beliefs in both Scotland and Ireland has long been recognised. In many respects, evidence for their early presence is there better represented than in England (Laing, 1977). For instance, extensive monumental evidence in the form of stone crosses and early Christian inscriptions (such as those detailed by Allen and Anderson, 1903), together with the historical record, provide confirmation of significant Christian activity in Scotland from as early as the fifth and sixth centuries. As centres of early Christian influence, the historic claims of Iona and Whithorn certainly appear to have been as important as those of Canterbury and Lindisfarne in England. Irrespective of this situation, there had been little indication or recognition of early ecclesiastical stone buildings to house and support this Scottish religious activity.

3.2 Available rock types in Scotland It has been emphasized that Anglo-Saxon builders in England relied extensively on very local stone for the construction of their churches (See sections 2.2.1 and 2.11). If, at that time, their counterparts in Scotland worked similarly with local materials, generally the rocks available for selection would have been restricted to indurated and hard Palaeozoic and Precambrian sediments, igneous rocks like basalts and granites, or metamorphic slates, schists and gneisses. These rocks contrast radically with the rocks available for AngloSaxon builders in England, where typically, relatively soft limestones and sandstones were employed (See sections 2.5 and 2.11).

Naturally exposed rock is, and was, more prevalent than natural forest in Scotland and Ireland, and evidence of early stone churches would be expected in these countries at least as frequently as in England. Although in Northumberland, peripheral to Scotland, Anglo-Saxon churches are not uncommon; in the lands to the north of the current border, in a region once falling within the bounds of ancient Northumbria, identified religious buildings of comparable age were considered to be absent. Cruden (1986, 24), for instance, wrote as follows of Anglo-Saxon influences: ‘… it would be unwise to think that there was any appreciable appearance of them in Scotland generally’; and ‘Until the end of the eleventh century Scotland was a disunited Celtic and tribal country’. In 2002, Fawcett provided a comprehensive review of Scottish churches but he identified no ecclesiastical structure of a date prior to the Norman Conquest of England.

English rock types of comparable hardness to those characteristically available to the early Scottish mason are materials like flint, Bunter quartzites and silcretes. Such intractable rocks when used by the Anglo-Saxon masons were left in an unaltered state, or at best only roughly broken (Potter, 1998; 2002; and see section 2.11). Although Scotland possesses a plentiful supply of potential building stones, to the early mason most of these materials were difficult to work. Without suitably compliant local stone, complex architectural features were unlikely to be constructed and they would not, therefore, be present. Pilaster-strips, double-splayed windows and long and short work would not normally be created in such circumstances.

With clearly identifiable Anglo-Saxon architectural features such as double-splayed windows, pilaster-strips and long and short quoining rarely, if ever, evident in Scotland; it is understandable that the majority of authors should have believed that most, if not all, existing Scottish ecclesiastical buildings in the country were of post, English Norman Conquest age. As in England, Norman influence in Scotland was accompanied with some level of written records of church construction and patronage. These records were far superior to those of earlier times, and this again must have helped to suggest that no church building activity had occurred prior to the period of Norman activity.

3.3 A review of early Scottish churches 3.3.1 Introduction Evidence resulting from the comprehensive and continuing studies of the stonework of the early churches of England, indicates that the Anglo-Saxon partiality towards stones that were placed with bedding orientated vertically was widespread. The apparent abrupt disappearance to the north of the border with Scotland, a boundary that had not been static, needed reconciliation. Recognising the difficulties related to the differences in available rock types, early Scottish churches were examined, therefore, to determine if, from their remains, workmanship adopting the same styles of vertical orientation once existed, perhaps in similar abundance to those in England. This would utilize the new evidence of the ‘Patterned’ style which might be seen in the quoins

The Scottish situation with regard to an apparent absence of early stone churches presented an ecclesiastical enigma. The author’s studies involving the earliest stone churches in England had revealed that a number of 56

CHAPTER THREE and arch jambs of the earliest churches: the absence of pilaster-strips in these buildings having already been noted by others (as Taylor and Taylor, 1965, 6).

burial-grounds and mausoleums, in particular for the wealthy (see, Spicer, 2000). In numerous instances, walls, doorways, windows and other structural features have been altered or supplemented to meet the requirements of individual burial and commemorative needs. This custom is almost unknown in England. Only where the church site had probably already been redundant prior to the Reformation does the site in Scotland remain uncluttered from these adaptations.

Certain aspects of such a study were expected to be more complex than the corresponding studies in England, again due to rock differences. Many of the local sedimentary rock types, like Palaeozoic greywackes or sandstones, in Scotland, are too fine grained or massively bedded to display obvious bedding planes. Experience in England had indicated that in those rocks where bedding planes were difficult to elucidate because of the rock composition and structure (rather than for reasons related to extraneous surface covers, distance or insufficient light), the Anglo-Saxon mason would have experienced the same complications and may have set the stones in an unintentionally haphazard orientation pattern. Put another way, the mason of the period could no more orientate any individual stone than could the modern researcher armed with a powerful hand lens. Such rare English instances could only be suspected on the evidence of other identifiable Anglo-Saxon architectural features elsewhere in the church fabric, such features that would be absent in Scotland. The Scottish difficulties would be further increased where structural features like quoins were constructed of igneous or metamorphic rocks, neither of which possesses bedding planes (although on occasions they may show a lineation).

Probably, because of the ready availability of suitable stone, it is far more common in Scotland to observe buildings in which quoins, windows and doors are depicted with stone of a contrasting colour to the main wall fabric (sometimes referred to as ‘rustication’). The framing stone is typically a more easily shaped and moderately fissile rock, such as the Devonian, Old Red Sandstone or Permian, New Red Sandstone. In Scotland, both unfortunately and unusually, about the mid-17th century, and apparently particularly prevalent in the 1730 to 1870 period, some masons chose to insert these contrasting stones in side-alternate fashion with, in many instances, the bedding planes set vertically, that is, they re-adopted the Anglo-Saxon or ‘Patterned’ style. This type of workmanship may be observed at Edrom Church (NT 827 558) in the reconstructed south-west quoin to the nave and more especially in the hearse/coach-house in the cemetery (both probably constructed c.1732). Hoddomcross Church (NY 178 735) largely destroyed by fire in 1975 and built it is believed in 1817; the remarkable ruined stone-built laundry at Abbotrule (NT 610 127), probably constructed around 1810; and the pump room at Strathpeffer (NH 486 583), built as late as 1892 (Figure 3.1), provide further examples. Fortunately, the time gap between the ‘Patterned’ style emplacement of stone and the later copy is sufficient generally for the two periods of workmanship to be distinguished.

This study, therefore, was initially intended to be explorative. It was considered quite unnecessary to examine every church in Scotland to determine if early specialist stone orientations would reveal previously unrecognised traces of the ‘Patterned’ styles. The geographical regions examined tended to be selected partly in accordance with local geology; in particular, areas were chosen where sedimentary rocks were likely to display bedding planes. It was also considered probable that the opportunity of discovering these early churches would be increased if the church fabrics on ruined sites were scrutinized, for in general such sites had been less extensively studied: furthermore, churches in use were more difficult to access. Potential sites for examination (in excess of 300 have been viewed) were in many instances noted following a scrutiny of Ordnance Survey maps.

Difficulties can arise, however, on ruined early sites where mausoleums (or other extensions) have subsequently been erected in the 17th to 19th centuries. Such an issue occurs, for example, at Ayton old church (NT 927 609): there the internal quoins to an 18th century north aisle are constructed, in part, of vertically emplaced and similar stone to that in the lower portion of the southwest quoin. The occurrence and similarities create a level of reticence in the pronunciation of ‘Patterned’ style origins to this church.

3.3.2 Further Scottish difficulties Early into the examination of Scottish church sites additional difficulties to those detailed in section 3.3.1 became apparent. The very extensive destruction which had been incurred in the history of these ruined ecclesiastical sites had not been expected. This material destruction of the building fabric did not, however, generally originate from Calvinist practices at the time of the Reformation, nor from the malevolence of English or other invaders (see McRoberts, 1968). Rather, the paucity of evidence of workmanship of early Christian origin in ecclesiastical buildings has been extensively determined by the post-Reformation practice of permitting vacated church sites to become personalized

3.3.3 Churches created in the ‘Patterned’ style Following an initial trawl of about 250 redundant or ruined Scottish ecclesiastical sites, 23, or nearly ten per cent, provided clear evidence from their stonework of having been built in the ‘Patterned’ style; that is, in a style comparable with that of the Anglo-Saxon masons in England. Further details are provided in Potter (2006d). It should be emphasized that this return does not reflect the extent to which the style occurs in churches which remain in use, as it tends to in England. Nevertheless, stone 57

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND Figure 3.1 Strathpeffer Pump Room (NH 486 583) was built as recently as 1892 and the walls were constructed of rock which is presumed to be dark, Devonian, Middle Old Red Sandstone. The stones of the quoins were placed in side-alternate style with the bedding orientated vertically in each stone (BVFR or BVFL). This style of construction was followed by some masons in Scotland, particularly in the 1730 to 1870 period, and was virtually a copy of the much earlier ‘Patterned’ style (see section 3.3.2). BH tie stones, however, tend to be absent in the more recent style. Note that the stones have been both cut and dressed. emplacement styles in Scotland certainly indicate the presence of stone church construction prior to the mideleventh century. Examples of this workmanship are illustrated in Figures 3.2 to 3.8. Each of these examples has been described more extensively in Potter (2006d). Figures 3.9 and 3.10 display features associated with one of the churches also referred to in Potter (2006d). On the Scottish mainland, on the evidence of the limited number of churches examined, only sandstones, from the Devonian (Old Red Sandstone), the Lower Carboniferous or the Permian (New Red Sandstone) proved sufficiently suitable to create vertically emplaced stones for early quoins or arch jambs, to thus meet the requirements of the ‘Patterned’ style. In four Hebridean sites the rocks selected were more unusual (Potter, 2006d, Table 1).

Figure 3.2 The south-west quoin of St Serf, Kirkton, Burntisland (NT 232 864), is built of relatively large blocks of Lower Carboniferous sandstone set in side-alternate fashion. Many of the stones are placed with their bedding orientated vertically in ‘Patterned’ style. From the ground upwards the orientations of most stones can be read with a hand lens as follows: stone 1, ?; BVFL; stone 3, ?; BH; BH; BVFL; BVFR; BH. 58

CHAPTER THREE

Figure 3.3 The south-west quoin of St Mary, Rothesay, Bute (NS 085 636). This quoin has been rebuilt at the base but from stone 5 and above (stone 9 has been replaced) the orientations in the Upper Devonian sandstone blocks exhibit a ‘Patterned’ style.

Figure 3.4 The ruined church of Little Dalton (NY 090 747) was briefly described by the Royal Commission for Historical Monuments for Scotland (RCAHMS, 1920, 42) who described the church as ‘probably early 16th century’. The church is here viewed from the south-east and the east wall (right) is clearly of different and much larger stone composition than the south wall. Large blocks of Permian create the south-east quoin, which like the north-east quoin, contains stones set in ‘Patterned’ style. 59

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 3.5 Old St Andrew church, Gullane (NT 480 827) provides a complex, much altered and ruined church site. The south doorway to the original nave, although somewhat modified, still shows stones of Lower Carboniferous sandstone exhibiting BVFIA orientation in ‘Patterned’ style; here shown in the east jamb.

Figure 3.6 The small, single-celled ruined church at Glenearn (NO 107 164) is built of Upper Devonian sandstone blocks. Each of the four principal quoins displays vertically orientated stones set in ‘Patterned’ style. The view is towards the north-east quoin; the east wall stands on a rough square plinth, and the church appears to stand on early foundations.

Figure 3.7 The ruined church of St Columba, Uidh, Eye peninsula, Lewis (NB 485 323), viewed from the south-east. The original church was apparently single-celled, but a large slightly narrower west annexe has been added (it has been suggested in the late 15th or early 16th century; RCAHMS, 1928, 13). The original church is built of local Lewisian gneiss beach boulders packed with small broken fragments of gneiss, extensive repairs having been completed with Permian, red sandstone, from the Stornaway ‘Beds’. The eastern quoins both include a number of blocks of gneiss which exhibit vertical orientation. 60

CHAPTER THREE

Figure 3.8 On the south wall of St Columba, Uidh, a 13th-14th century doorway remains. To its left (west), the west jamb of an earlier infilled doorway is visible, this doorway has some evidence of ‘Patterned’ workmanship.

Figure 3.9 The early Christian cross at Kildalton church, Islay (NR 458 507), viewed from the south-west. Although modified, the four principal quoins at Kildalton, in their lowest stones, show some evidence of ‘Patterned’ emplacement.

Figure 3.10 This boulder in the north-west wall at Kildalton church shows Pholas borings, indicating is collection from the sea shore.

Prior to the study of the stonework in the Scottish sites, a few authors had provided arguments for offering a preEnglish Norman Conquest age for a restricted number of Scottish ecclesiastical sites. Taylor and Taylor (1965, 710-13) proposed that the tower of Restenneth (NO 483 516), the tower and chancel of St Rule, St Andrew’s (NO 515 167), and the excavated church beneath Dunfermline Abbey (NT 090 873) were of Anglo-Saxon origin. Fernie (1986) placed the two standing structures (Restenneth and St Andrew’s), together with four other early buildings

(Abernethy [NO 190 165], Brechin [NO 596 601], Egilsay [HY 466 304] and Edinburgh Castle chapel [NT253 735]) in the period 1090 to 1130. In response, Cameron (1994) rightly argued the case already discussed above, that such a late origin for these churches raised questions about the relative backwardness of the first stone-built churches in Scotland. At both Restenneth and St Andrew’s, Taylor and Taylor were able to distinguish architectural features which were the same as in English, Anglo-Saxon churches. Features of the ‘Patterned’ style 61

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Church

Patterned feature displayed

Where displayed Best in northern quoins

St Margaret, Edinburgh

Vertically bedded stonework in quoins Vertically bedded stonework in door jamb Colour banding Face-bedded pattern

St Peter, Restenneth

St Rule, St Andrews

Poor, in north jamb of blocked west doorway Three principal walls Poor, in certain walls

Vertically bedded stonework in quoins Vertically bedded stonework in door jambs Vertically bedded stonework in arch jambs Vertically bedded stonework in window jambs Cut back stonework in doorway architrave Cut back stonework in windows

West quoins of tower, at higher level all quoins South doorway (but altered), east doorway tower Tower east arch (much altered) Most high level tower windows

Double-splayed windows Vertically bedded stonework in quoins Vertically bedded stonework in arch jambs Face-bedded pattern

Chancel

South doorway (incomplete) Triangular-headed windows

Each of four chancel quoins, east nave quoins Outer order of jambs of eastern arches (poor) Chancel walls in particular

Table 3.1 ‘Patterned’ stonework features recognised in three early churches in Scotland

Figure 3.11 The site of the priory church of St Peter, Restenneth (NO 482 516) viewed from the north-west. The roof line of the nave is clearly marked on the tower. The quoins of the tower are set in ‘Patterned’ style (Potter, 2008c).

Figure 3.12 The east side of the tower arch at Restenneth church, the jamb stones of which, although now somewhat modified, show evidence of ‘Patterned’ insertion. 62

CHAPTER THREE

Figure 3.14 The south-west quoin of the chancel of St Rule, St Andrew’s is set with its stones in ‘Patterned’ style. From the plinth the stones read upwards: BH, BVFL, BVFR, BVFL, BVFR. In the first four courses, the clearly visible stones in the south wall are, with the exception of BVFL quoin stones, face bedded.

Figure 3.13 St Rule, St Andrew’s (NO 515 167) is constructed of a Lower Carboniferous fine grained sandstone. The church has recently been described by Potter (2008c) and is here viewed from the north-west.

Figure 3.16 The south wall of the chapel of St Margaret, Edinburgh Castle (NT 253 735) reveals evidence of four styles of building at different levels, one of which may represent ‘Patterned’ colour banding (see Potter, 2008c). Figure 3.15 The western tower arch of St Rule, St Andrew’s was rebuilt and subsequently blocked. 63

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 3.18 Abernethy round tower (NO 190 165) is here viewed from the north. The change in stone type in the tower (referred to in section 3.3.4) may be observed to the right of the doorway.

Figure 3.17 Detail of stones which are believed to have been inserted in ‘Patterned’ style in the north-west quoin of the Edinburgh Castle chapel. identified in one or other of these two churches included: double-splayed windows, triangular-headed windows, plain square pilaster-strip and hoodmoulding surrounding arches, through stones in arches, simple, plain square plinths and string-courses, square unbuttressed towers and (p. 712) double belfry windows ‘similar to those of Anglo-Saxon towers’. Each of these ecclesiastical sites has been examined, but to differing extents, by the present author. The foundations of a two-celled church and an apsidal extension beneath the Norman nave of Holy Trinity, Dunfermline can only be viewed by windows in the nave floor, so that detail of the stonework cannot be determined. The important facets of the ecclesiastical buildings at Restenneth, St Andrew’s and Edinburgh Castle, all being national monuments, have been examined and described more specifically (Potter, 2008c). Some supplementary photographs of these three church sites are provided in Figures 3.11 to 3.17.

referred to by Fernie (1986) is briefly considered below (see sections 3.3.4 to 3.3.6). 3.3.4 Abernethy Round Tower, Perth and Kinross (NO 190 165) The round tower at Abernethy is today Scotland’s only wholly free-standing tower (Figure 3.18). Both the date of its foundation and its possible associations with the towers of Ireland, have been matters of much conjecture. Cruden (1986, 6 and 10) believed the lower parts of the tower to date from about 1000. Fernie (1986) assigned a date of around 1100 or later to the building. That, in 1089, King Malcolm of Scotland paid fealty to William the Conqueror at Abernethy has indicated to many the early importance of the site, and unsubstantiated records of 9th to 12th century (McNeill, 1974, 208) and 11th century (Laing, 1977, 71) indicate the range of possible suggested dates of origin. The presence of an early 7th century Pictish inscribed stone at the base of the tower further complicates this issue of precisely when the tower was erected. Cameron (1994) noted the masonry break which occurs low in the tower (Figure 3.18) which implied considerable rebuilding. This break was also noted by Brash (1857-60) in his remarkably accurate and detailed description of the tower. The following observations relate only to a brief internal ascent of the tower and limited investigation.

Distinctive ‘Patterned’ style stonework features included in the descriptions of the three monuments were; vertically bedded stones in quoins and arch jambs including windows, wall colour banding, cut back architraves, and patterned, face and edge bedded wall surfaces (see also Table 3.1). These characteristics, together with the architectural detail, are believed to be sufficient to place all three buildings in a period prior to that which, in England, would have been termed ‘Norman’ or ‘Romanesque’. Each of the other three sites 64

CHAPTER THREE sandstone lithologies). To the left (east) of the doorway, the (slightly wider) upper stone is now placed BVEIA. It carries the marks of what might be interpreted as a worn carved figure. This ‘figure’ is noticeably not upright within the stone and it probably represents unusual weathering of the softer sandstone (Figure 3.19). Brash (1857-60, Plate XXXII; herein Figure 3.20), provided an elevation of the doorway, on this the jamb stones were all cut back in ‘Patterned’ style. The ‘figure’ could represent some trace of this cut back. That the Brash drawings are likely to provide an accurate representation is suggested by the right (west) jamb. His illustration depicts only two stones (both of which are set BVFIA); today, the top of the lower stone has been partially replaced (with a BVEIA stone). The doorway tapers upwards in ‘Patterned’ style. Brash commented on the enormity of both the doorway and the outer over-arching stone. His records for the doorway can be converted to the following metric measurements; height 2.34m, width at sill 0.80m, width at springing 0.72m. As far as can be determined, with the help of binoculars, the exterior of the four large ‘belfry’ windows to the tower are each of Anglo-Norman, Romanesque style (Cameron, 1994). However, internally, the orientation of their jamb stones suggests an earlier origin. Three of the windows have been altered to varying degrees, and replaced stones, absent stones and cemented areas are present. The window on the south wall is the best preserved and its jamb stones (four on each side) display the following bedding orientations:

Figure 3.19 The doorway to Abernethy round tower, in which all the stones were originally placed in ‘Patterned’ style.

East jamb detail (left as viewed) The stonework in the lower, visibly examinable, portion of the tower is complex. Three different lithologies of sandstone are involved. They are probably all of Upper Devonian/Lower Carboniferous age, but absence of local natural rock exposure makes even this statement difficult to confirm. Brash (1857-60, 310) wrote, ‘Tradition states that the building materials were brought from the Lomond Hills’ (about 12km. to the south), but then continues (!), ‘…they were conveyed by a line of men … the stones being handed from man to man’. The sandstones are; a yellowish, medium and uniform grained, somewhat softer variety which occurs above and surrounds most of the doorway; a buff, slightly harder even grained sandstone which appears to have been used for repairs; and a greyer, harder, less well sorted sandstone in which feldspar occurs and muscovite mica is more prevalent. This sandstone can be observed principally forming the wall low on the west (immediate road) side of the tower (Figure 3.18) but it makes up the major part of the stone in the tower base below the sill of the doorway and must be considered as the original stump of an earlier tower.

BVFIA (through stone) BVFIA BH (through stone) BVFIA

West jamb detail (right as viewed) BVFIA (through stone) BVFIA BH (through stone) BVFIA

It is possible that this work could have been completed subsequent to the mid-17th century (see section 3.3.2 and Potter, 2006d, 231), but the likelihood must be that it is of ‘Patterned’ style. This would support the evidence from the jambs of the doorway. Without more extensive investigation the author can only currently assume that the tower was both built and rebuilt during the preNorman, ‘Patterned’ period. 3.3.5 Brechin Cathedral Round Tower, Angus (NO 596 601) It is normally believed that Brechin round tower remained detached from the cathedral until the 13th century when the cathedral was rebuilt. Like Abernethy, its age is controversial. The ornate doorway (Figure 3.21) has been extensively discussed and it has been this feature that has provided the principal arguments as to the tower’s date of origin. Although, Cameron (1994) presented documented reasons for tracing the history

The jambs to the doorway show evidence of modification. Each jamb probably originally consisted of two vertically emplaced BVFIA stones (of the first of the 65

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 3.20 The illustrations of Abernethy round tower as produced by Brash (1857-60, Plate XXXII). The doorway jambs as drawn should be compared with those in the recent photograph.

Figure 3.21 The doorway to the round tower at Brechin Cathedral (NO 596 601) possesses all the characteristics of being constructed in ‘Patterned’ style. The stones in the doorway are set BVFIA and exhibit cut backs.

Figure 3.22 The illustrations to the paper by Jervise (1857-60) of the round tower at Brechin. There are errors in the drawing of the doorway.

66

CHAPTER THREE back to about 1017, Fernie (1986) made Irish comparisons to place the tower in the 12th century. Various other authors have proposed dates generally falling within the range of these two authors. An early paper by Jervise (1857-60) fails to provide the same descriptive precision or accuracy (Figure 3.22) as that of Brash (1857-60) for Abernethy.

The stones of the doorway and the tower are created from the local Lower Devonian sandstone. The stones used in the tower walls are large and set BH and, except where repaired, the walls do not show the occasional insertion of edge bedded stones typical of Norman Romanesque work (see section 2.12.3). 3.3.6 St Magnus, Egilsay (HY 466 304)

Although the present author has only briefly examined the stonework of the exterior of the tower from ground level, certain distinctive features of the ‘Patterned’ style of construction are immediately evident in the doorway, thus supporting the views of Cameron (1994) as to the date of the tower’s origin. A single megalithic stone forms the jamb on either side of the doorway which tapers slightly upwards towards the single megalithic arching stone. Each of the jamb stones is placed BVFIA and is slightly cut back on its outer edge. The arching stone, set BH, is also cut back to carry the proud face width of the jamb stones in the form of an architrave. Also cut out of the arching stone are the two ‘slabs’ set as ‘shoulders’ to the arch; these ‘slabs’ Jervise (and his artist Ormiston; Figure, 3.22) incorrectly placed outside, and independent of, the arch. The carved BH lintel stone is also cut back to carry the proud face width of the jamb stones. It is stated (although not observed by the present author) that the two jamb stones are through stones. All the ornate figures and carvings have been skilfully created on the face of the four large stones which make up the frame of the doorway; the whole structure of which is distinctively ‘Patterned’ in character.

The ruined church on the island of Egilsay, Orkney, provides very significant problems with regard to the interpretation of both its architecture and its remaining fabric. For these reasons the following descriptions and analyses will be presented in rather more detail than provided in all previous sites referred to in this monograph. Particularly in architectural terms, a variety of matters will be shown to still require further widespread study and elucidation. Some of these will be summarised in Chapter 7. A plan of the church is provided (Figure 3.23). In historic terms the island of Egilsay is famous for the Martyrdom of St Magnus Erlendson, supposedly about 1115 (Pålsson and Edwards, 1981, 96, footnote). The Royal Commission (RCAHMS, 1946, 3, 229) provided an exact date as 16th April 1116. The Earldom of Orkney had been shared by Magnus and his cousin Haakon for twelve years, but Haakon is said to have become jealous of the popularity of Magnus, and either Haakon, or probably one or more of his followers, killed Magnus on the island. Magnus is recorded (although versions vary)

Figure 3.23 A plan of St Magnus Church, Egilsay. Key: A1 – south doorway; A2 – north blocked doorway; B1, B2 – early single-splay blocked windows to nave; C1, C2 – early single-splay blocked windows to chancel; D – access doorway to tower (with nearly identical doorway on first floor above); E – above chancel arch, doorway nave first floor to chancel first floor, with step; F- first floor early window to tower; G – later rectangular window to chancel first floor; H1 – late rectangular window above south nave doorway (preserves wooden frame); H2 – late blocked rectangular window (preserves part of wooden frame); H3 – late blocked and altered, rectangular window ground floor tower; J – trace of original narrower chancel arch.

67

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 3.24 The problematical church of Egilsay, Orkney (HY 466 304) as viewed from the south-east. as using a church for prayer on the evening immediately prior to his murder, but whether the existing church ruins are those in which he sought sanctity is not known (Ritchie, 1993, 111).

1992, 297); the Royal Commission simply proposed ‘12th-century’. All three principal parts of the church are roofless today but the chancel still preserves a barrel-vaulted ground floor. Hoey and Thurlby (2004, 166; citing Fernie, 1988) noted this occurrence, and included its construction within the date ‘1066 to around 1170’ (p. 117).

The church is a simple three-celled structure comprising, on the same axis, a tower, nave and chancel (Figure 3.24). Although the tapering tower, observed from a distance, resembles those seen in Ireland, it is adjoined to the ‘west’ wall of the nave in a manner more closely allied to the style of the round towers of East Anglia (Fernie, 1986, 394). In this respect the church is unique in Scotland. The presence of the tower, however, has caused some to relate the church to others at Abernethy and Brechin, collectively attributing the three with the possession of Irish associations (Marwick, 1952, 32; Ritchie and Ritchie, 1978, 68).

Analysis of the church fabric has been additionally complicated by a number of factors. The island of Egilsay possesses, apart from deposits, like till, resulting from glaciation, only one, very uniform, rock type. All the rock used in the church, whether in any original building or repair work, has been extracted from the island and, is virtually identical. It can be identified as rock from the Middle Devonian, Rousay Flags Sub-Group. These greygreen flagstones vary slightly in grain size, and occasionally, as may be seen in the tower, they may include small areas of chert. Repair work, therefore, proves nearly indistinguishable and little more than block and rubble size and shape assist in structural interpretation. No assistance has been afforded by the repointing of the whole church, probably on more than one occasion, in an identical cement/lime-based mortar. Although they acknowledged an earlier ‘Celtic church’ on the site, Dietrichson and Meyer (1906, 23-24) argued, but with very little reference to the building fabric, that the present church was Norwegian in character (Figure 3.25). If this were to be true a detailed knowledge and comparison with early Nordic churches would be necessary to confirm the view.

Dates for the construction of the church vary significantly, although the majority view is that the present structure dates from shortly after the death of St Magnus (RCAHMS, 1946, 228-229), so that the building commencement is often placed more precisely to the year 1116 or 1117 (as Fernie, 1986; Fawcett, 2002, 26). Cruden (1986, 10) believed the church dated from the late 10th or early 11th century. The most extensive studies of the church are probably those of Dietrichson and Meyer (1906, in Norwegian, with English summary, 23-28) and RCAHMS (1946). The former authors reasoned from literature records that the church was erected between December, 1135, and 1138 (a date supported by Gifford,

68

CHAPTER THREE

Figure 3.25 The nave and tower of Egilsay church viewed from the south and drawn by Meyer for publication in Dietrichson and Meyer (1906). Without repeating the various observations of other principal authors regarding the site, considerable supplementary detail can be added to the church description. The church is recorded as being in use in 1841 (The Statistical Accounts of Scotland, 1834-45, Vol. 15, 84). From this it must be inferred that, at the time, the nave, at least, must have possessed a roof (a roof to the upper level of the chancel would seem likely, but its absence would not necessarily have precluded the use of the church). This nave roof must relate to the currently visible crow-stepped (or corbie) gables (Figures 3.26 and 3.27). In the south wall of the nave (Figure 3.27) a rectangular window above the south door is a later insertion (Gifford, 1992, 297, suggests of 17th century date), which must have provided light to the upper levels of the roof. Access from the nave to both the tower and the chancel appears to occur at two levels; the ground and the first floor levels (Figures 3.28 and 3.29). The precise arrangement of the upper access points to the ‘first floor of the nave’, unfortunately, is lacking in evidence. Dietrichson and Meyer (1906, 23) suggested that a

gallery ran round the nave walls between the doorways of the tower and chancel below an ‘open timber roof’. The RCAHMS (1946, 228) refers only to ‘the roof space above the nave’, which if it were a room would leave the ground floor nave ceiling height restricted. There are no visible corbels and the small holes which are evident (at three levels) on the inside of all but the west wall are probably putlock holes. A wooden stairway or ladder from the nave to both access points would seem to have been unlikely, but the entry into the chancel is noticeably not at the same level as that into the tower. The rectangular window (Figure 3.27, which preserves some of its wooden frame) penetrates the nave wall at a level approximating to the ‘floor’ between the ground and first floors of the nave, and would, therefore, seem to postdate the presence of such a floor. It seems likely that, at least internally, the upper levels of the church, from the evidence provided from these relationships, were completely rebuilt prior to the insertion of this rectangular window.

69

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 3.26 The crow-stepped or corbie gable at the east end of the nave of Egisay church viewed from the south-west.

Figure 3.27 The corbie gable at the west end of the nave of Egilsay church, viewed from the south. The small rectangular window is referred to in the text. The crow-stepped gables to the nave seem to follow the style of many similar gables in both Orkney and neighbouring Caithness; for they are constructed in two layers with frequent tie stones: the corbie outer layer being about 35mm. thick and the supporting inner layer about 55mm. thick; providing a wall thickness of about 90mm (as Figure 3.26). A review of the more reliable literature for northern Scotland provided, from 60

instances of crow-stepped gables, dates for their erection of no earlier than about 1620. This would again suggest that the upper portions of Egilsay have been modified or rebuilt about this or some subsequent, date. Figures 3.28 and 3.29, illustrating the inside of each gable do show a marked change in stone size in their upper portions. No evidence of an earlier nave gable or roof line is visible, however, on the east wall of the tower. 70

CHAPTER THREE

Figure 3.30 The blocked north doorway in the north nave wall of Egilsay church.

Figure 3.28 A view of the interior of the west end of the Egilsay nave to show the relationship with the tower and the access entrances to the tower at ground and first floor levels.

Figure 3.29 A view of the interior of the east end of the nave of Egilsay church, showing the chancel arch, the simple doorway, with step, at first floor level and, the corbie or crow-stepped gable.

Figure 3.31 The blocked north window in the north wall of Egilsay church has been infilled in the same style as the doorway. 71

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 3.32 Each of the jambs of the chancel arch show evidence of having been partially destroyed to widen the arch. The south jamb is here viewed from the north-west. It can be noted that the base of the jamb projects north of the chancel wall. The barrel-vaulted chancel originally possessed narrow, single-splayed windows in the north and south walls. These are now infilled. Of the doorway and windows on the ground floor that have been infilled (Figure 3.23; A2, B1, B2, C1, C2, H2 and H3) all have been blocked in the same manner (Figures 3.30 and 3.31), suggesting that this was done sometime after the church fell into disuse. In the upper storey, there is a small, rectangular, later window in the east wall. The chancel arch displays evidence from its partially destroyed jambs of having once been narrower (Figure 3.32).

would have proved difficult, set BH, to insert in the walls, for they would have been too big to fit securely and comfortably with other stones. There is some visual evidence that, like the end walls of the nave, all the major walls of the church may be ‘double-skinned’, further adding to the intricacies of using exceptionally large stone blocks. One particular face bedded block is worthy of possible further examination and from its shape it might be a re-used gravestone: it may just be observed from ground level, to the north of the nave-chancel upper door on the chancel west wall.

The fabric of the church nave and chancel walls displays a remarkable mix of large and small irregularly broken stones (Figures 3.33 and 3.34); the smaller fragments typically set BH, but most of the larger blocks being face bedded; a building style also observed in certain Irish churches (see section 5.2). The constituent rock is a hard, but brittle, flagstone which accounts for the size range in the building fabric. Attempts to break the large blocks would have resulted only in small pieces. At Egilsay, the mix of large and small fabric is evident both on the exterior and, where the wall plaster no longer exists, the interior of the walls. The largest of the face bedded stones reach exposed dimensions of in excess of one metre, typical sizes being; 1.25m. x 0.45m., 0.95m. x 0.55m. and 0.75m. x 0.60m. Most of these blocks are probably no thicker than 0.10m., as can be observed at the wall quoins. The question as to why these blocks were not set BH, as the smaller material, must be considered. With the exception of the west nave wall which reaches a thickness of 119mm., all other major walls are of the range 91 to 95mm. thick. Without breaking the large blocks, they

All six principal quoins include the larger face bedded stones and, noticeably, these are placed, in most instances, alternately in the quoins to the BVFR-BVFL pattern. In each quoin, however, numbers of small BH stones are interspersed in the ‘pattern’. The south-east chancel quoin, proceeding upwards from ground level, may be cited as an example (see Figure 3.35): stones 1-2 BH, 3 possibly BVFR, 4-6 BH, 7 BVFL, 8 BVFR, 9 BVFL, 10-12 BH, 13 BVFR, 14-15 BH, 16 BVFR, 17 ?BVFR, 18-19 BH, 20 BVFR, five replacement stones BH to roof line. The south-west nave quoin is illustrated and detailed in Figure 3.36. The other four major quoin details are cited below: North-east chancel quoin: stones 1-15 BH, 16 BVFR, 17 BH, 18 BVFR, 19-21 BH, 22 BVFL, 23-28 BH replacements to roof line. North-east nave quoin: stones 1-10 BH, 11 BVFR, 12-20 BH, 21 BVFR, 22-23 BH, 24 BVFR, 25-27 BH, 28 BVFR, 29 BVFL, 30-38 BH replacements.

72

CHAPTER THREE

Figure 3.33 View of the nave and tower of Egilsay church from the north. The fabric of large face bedded stones and smaller horizontally set rubble should be noted.

Figure 3.34 The south nave doorway of Egilsay church is here viewed from the south-east. The style of the stonework of the wall noticeably changes about a metre from the ground to include large face-bedded stones.

Figure 3.35 The south-east chancel quoin of Egilsay church has many stones set to the BVFR-BVFL ‘Patterned’ style. These stones are often interspersed with smaller BH stones. Large stones 7 to 9 are visible, see text for detail. 73

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 3.37 Stones below the infilled window in the south chancel wall of Egilsay church are clearly face bedded and one stone preserves fossil ripple marks.

Figure 3.36 The south-west nave quoin of Egilsay church. In this quoin the stones are placed as follows: stones 1-6 BH, 7 BVFR, 8-13 BH, 14 BVFL, 15-17 BH, 18 BVFR, 19-24 BH, etc. The larger stones 7, 14 and 18 can be distinguished. (The pen is 135mm. in length).

Views of the wall stone fabric overall do reveal significant changes in the stone size. Apart from the obvious wall capping areas, this is especially visible in the south nave wall and elsewhere, where the lowest metre or more is devoid of all large stones (Figure 3.34). With such a resistant stone in use the author is inclined to suggest tentatively that this represents the original wall stonework in a church that was once extensively ruined. Such a church may have possessed clearly ‘Patterned’ quoins and, inside its chancel, face bedded stones.

North-west nave quoin: stones 1-13 BH, 14 BVFL, 15 BH, 16 BVFR, 17 BVFR(repair), 18-27 BH, 28 BVFR, 29-33 BH, 34 BVFL, 35 BH, 36 BVFR, 37-38 BH, 39 BVFR, 40-43 BH replacements. South-east nave quoin: stones 1-11 BH, 12 BVFL, 13-20 BH, 21 BVFL, 22 BH, 23 BVFL, 24-26 BH, 27 BVFR, 28 BVFR, 29-30 BH, 31 BVFL, 32-?43 BH replacements. The flagstone material used in the church is both hard and durable and it is difficult to argue for certain that all the small BH stones in the quoins are replacements, and thus conclude that the workmanship was once of orthodox ‘Patterned’ style. It is, however, possible that in its early history the church fell into disuse and that quoin stones (particularly those closest to the ground) were dislodged intentionally for use elsewhere (as may be observed in nearly all long disused churches). It is also possible that unusually large stones were too big to place in BH attitude even in the quoins. None of the older doorways or windows possesses any clear evidence that they display the ‘Patterned’ style of stone insertion.

3.4 ‘Patterned’ churches in the Isle of Man 3.4.1 Introduction A number of early churches in the Isle of Man have been reviewed briefly. Just nine sites were visited and, of these, a number were early keeills now lacking any, or possessing very limited, visible stonework that had not been disturbed. These small structures (Figure 3.38), often associated with related early cross slabs (Figure 3.39), are rarely bigger than 5m. x 3m. internally, and are very similar to numerous Christian sites in Scotland (and Ireland). All the church sites examined professed largely unsubstantiated claims to an early origin, generally in the range of the 6th to the 8th centuries. Nearly all the standing buildings examined had been rebuilt, in a number of instances in the 12th century, so that clear evidence of ‘Patterned’ workmanship in them was absent.

Possibly the strongest suggestion of a ‘Patterned’ age for parts of the church may be seen inside the chancel. Low in the south wall the stonework is all set in a face bedded style. One of the stones preserves fossil ripple-marks, which particularly exemplifies this orientation (Figure 3.37). 74

CHAPTER THREE

Figure 3.38 The keeill, or small early chapel, known as Lag ny Keeilley (SC 216 746), which nestles in the cliffs near Niarbyl Bay, Isle of Man, serves as a local picnic site. Better preserved than most keeills its walls are estimated to only be of the order of 675mm. thick. One rock unit predominates in the Isle of Man: the Cambrian-Ordovician, Manx Slate Group. These rocks consist of a complex series of slates, greywackes, siltyflagstones, etc. (Simpson, 1963), all of which exhibit a limited element of metamorphism. As a result lineations imposed on the rocks, and particularly the finer grained slates, are sometimes the result of this metamorphism, and they may not, therefore, represent the initial depositional layering or bedding of the rocks. The island also has smaller areas of other rock types, including red Palaeozoic sandstones which have been used occasionally, in some cases for ornamental affect, in certain churches. Only one church, of those examined, showed clear indications of ‘Patterned’ craftsmanship and this is described briefly in Section 3.4.2. Instances of vertically orientated stones in structures like quoins in other churches, provided for varied reasons, unreliable detail (Figure 3.40). 3.4.2 St Patrick’s Isle, Peel St Patrick’s Isle is now linked to the west side of the mainland of the Isle of Man by a causeway. It possesses an interesting and extensive archaeological history, which includes a number of early Christian sites. Set in the grounds of the later Peel Castle, these include the ruined early 13th century cathedral, a detached round tower, the ruined church of St Patrick, a partially reconstructed keeill, and a further, no longer visible, recently excavated chapel (of 10th-11th century, Freke,1995).

Figure 3.39 One of many early Christian Celtic crosses in the Isle of Man. This, partly repaired cross occurs in the churchyard of St Adamnan, Lonan (SC 426 794). Apart from an early plinth and large worn boulders at the base of certain walls no trace of an early church is really present at the church site.

The single-celled, ruined church of St Patrick (SC 242 847) has normally been regarded, like the round tower, as dating from the 10th or 11th century. The church reveals 75

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND three different features of ‘Patterned’ stonework. Each of the four principal quoins contains two vertically orientated stones at, or just above, its base. The north-east quoin, for instance, stands on a simple early plinth (BH), with BVFR, BVFL, orientated stones immediately above. Higher stones in the quoins are set BH. In the south wall of the church the three lowest courses of local red, ?Devonian, Peel Sandstone are laid in ‘Patterned’ style: Row 1, BH; Row 2, mainly face bedded; Row 3, BH. Higher, across the wall, the rocks are initially all placed in herringbone fashion (Figures 3.41 and 3.42). The herringbone work is here normally described as being of 12th century age. In the west portion of the south wall, the church once had a south door: this had ceased to be used and was infilled prior to the creation of the herringbone coursework. Clearly, the church had both been built and fallen into ruin (or been extensively destroyed) prior to the herringbone work being commenced. The round tower immediately to the west of the church has signs of significant rebuilding. Although the doorway narrows towards its arch, the detail of the jamb stones could not be scrutinized from ground level to assist in determining the tower’s possible age (Figure 3.43). A suspicion remained that any early work had been obliterated in the rebuilding. St Patrick’s keeill or chapel to the north of the church of St Patrick, is claimed to have been founded in the 10th century. It possesses a west doorway, but the ruined site similarly shows no evidence of ‘Patterned’ workmanship, with some of the additions being even late medieval in age.

Figure 3.40 These three stones, set BVFL, BH, BVFL, occur as part of a short quoin at the base of the northeast corner of the north transept of the Cathedral of St German, Peel (SC 242 846). The cathedral was not built until the 13th century and it is difficult to relate either the quoin or its short length of adjoining wall to the cathedral or other adjacent earlier buildings (lens cap 50mm in diameter).

Figure 3.41 The ruined church of St Patrick, St Patrick’s Isle, Peel (SC 242 847) viewed from the south-west. A summoner’s oratory or pulpit stands on the west side of the nave. 76

CHAPTER THREE

Figure 3.42 Detail of the south wall of the nave of the church of St Patrick, Peel. The herringbone work in the wall succeeds several courses of Peel sandstone blocks laid in ‘Patterned’ style. See text for detail.

Figure 3.43 Viewed here from the east, the detached round tower to the west of the church of St Patrick, Peel, shows considerable evidence of alteration. 77

CHAPTER FOUR A GEOLOGICAL REVIEW OF MASONRY STYLES IN IRISH EARLY ECCLESIASTICAL ARCHITECTURE

4.1 Introduction ‘There has also been considerable controversy concerning the origins, dating and chronology of early medieval stone churches in Ireland because of their architectural simplicity.’ Edwards (1996, 122)

Studies relating to early ecclesiastical buildings and their dating and construction normally rely heavily upon the examination of recognisable and, what develop with experience into, dateable architectural features. Such studies are supported by any available historical documentation, and less frequently with archaeological excavations. Irish church studies have pursued these techniques comprehensively. The work of Taylor and Taylor (1965), in England, did much towards emphasizing the importance of recognising and interpreting cross-cutting architectural relationships, and there are now many occasions where structural interpretation of this nature has been applied effectively in Ireland.

Most authors concur with such views: for instance, Leask (1955, 1) wrote ‘the precise dating of any work of primitive building is difficult and often impossible’. For many authors in Ireland, the early ecclesiastical building prior to the onset of the Norman inspired ‘Romanesque’ period was principally in wood (Harbison, 1982; Hamlin, 1984; Hare and Hamlin, 1986). Somewhat earlier, McNeill (1974, 131) had stated, ‘in architecture, Ireland remained a backward nation until the Anglo-Norman invasion’. O’Keeffe (2003, 16) presented a somewhat more positive analysis for he noted, as had others, that the introduction of Gaelic or Irish Romanesque (HibernoRomanesque) building activity and styles had occurred well before the beginnings of the Anglo-Norman invasion in 1169. That is, as in Scotland, the ‘Romanesque’ fashions in building preceded any influx of the Norman peoples.

Although ecclesiastical geology is currently in its infancy, it has proved of significant value in England, where structures erected by Anglo-Saxon church builders may now be distinguished with increased accuracy. In Scotland and the Isle of Man a new ‘Patterned’ style of workmanship has been revealed. In this Chapter, the two main tenets of ecclesiastical geology, the detail of the stone and the manner in which it was used, will be applied to Irish churches.

‘…the fact that Gaelic Ireland had adopted this evidently multi-national medium entirely of its own volition, without even the mediating presence of a colonial population on the island …as extraordinarily interesting from the perspective of cultural history.’ O’Keeffe (2003, 16)

As in Scotland and the Isle of Man, the historical records, together with extensive evidence of early Christian artifacts, such as monumental cross slabs, high crosses and certain Ogham inscriptions, provide support for extending the origins of Christianity in Ireland back to dates well prior to those typically applied to the earliest stone churches. It has been suggested that Christianity was introduced into Ireland as early as the 4th century (as Hurley, 1982, 297). Such suggestions are based largely upon hagiographic records many of which are probably legendary (McNeill, 1974, 50). By the latter part of the 6th century Irish monasticism was flourishing (McNeill, 1974, 69 et seq.), and many important monasteries, such as those at Clonmacnoise, Durrow and Bangor in Co. Down (Edwards, 1996, 99) were founded. From these and other monasteries, remarkable illuminated manuscripts such as the Book of Durrow (c. 650) and the Book of Kells (c. 800), and the various, historically informative and much cited, Annals, ultimately resulted.

Popular today is the view that, provoked possibly by Viking raids (Leask, 1955, 2), some stone churches were constructed in Ireland from the 9th and 10th centuries. Documentary sources suggest that prior to the 9th century, and then possibly only from the 8th century, a few major ecclesiastical sites, such as Duleek and Armagh were built in stone (Harbison, 1982, 620; Hamlin, 1984, 118; Edwards, 1996, 131). It is only in the 11th and 12th centuries that references to stone churches become more frequent. Irish early stone churches lack most of the English forms of ‘Anglo-Saxon’ variety or architectural decoration. This situation is only slightly inferior to, but not unlike, that in Scotland. Decorative features like pilaster-strips, so dependent on suitable workable stone, are again absent. Double-splayed windows are thought to be absent and there is nothing analogous to the double belfry tower window and its well-turned mid-wall shaft: all aspects of a reliance on suitable stone. Hare and Hamlin (1986, 133) itemized further deficiencies in the architecture and architectural sculpture of the early Irish stone church.

Despite this background, little documentary data remains to assist in the dating of early Irish churches. Hare and Hamlin (1986, 131) went as far as stating that ‘…there is no single pre-Romanesque church in Ireland to which a firm date can be assigned’. This statement was offered in spite of ‘an estimated two thousand early ecclesiastical sites …’ having been identified already in the Irish landscape (Swan, 1994). The simplicity of most of the early churches provides additional difficulties. 78

CHAPTER FOUR Recognising that because, in addition, there is a paucity of documentary evidence creating difficulties in the dating of early Irish churches; archaeologists have examined alternative means of determining their possible age. Small churches and oratories of drystone construction occur in the area of south-west Ireland and to some their lack of mortar was interpreted as evidence of an early origin. This caused Leask (1955, 1) to write, ‘If the date of the introduction of the use of mortar were known the task (of precise dating) might be made easier’. With similar corbel built structures (beehive huts) being built in very recent times the absence or occurrence of mortar would seem to provide little evidence of age. Harbison (1982) divided the early churches in Ireland into four major varieties: those built in the corbelled technique, simple rectangular unicelled structures, similar forms with antae (see Chapter 6), and structures with both nave and simple chancel; he then attempted to offer tenuous dates for the occurrence of each variety. Others, such as Ní Ghabhláin (1995, 94-105) and Ó Carragáin (2005a) have sort to utilize wall masonry styles for the same purpose of church dating. A particular aspect of wall masonry, the inclusion of cyclopean or megalithic stones, was used by Petrie (1845) considerably more than a century earlier, to indicate that in his belief these buildings had great antiquity. Manning (1995a, 16) has interpreted this type of stonework as probably distinctive of churches of the late tenth and 11th centuries, he has also suggested (Manning, 1998, 76) that antae which project more extensively from the church end walls are indicative of a relatively early date. Various authors have proposed a number of additional age-dating estimates based on other different aspects of the fabric of both the churches and their associated round towers.

Irish geology provides only limited areas of stone that can be worked with ease into ashlar, dressed or sculptured blocks. Hard and durable, Palaeozoic sediments, igneous and metamorphic rocks predominate; the more readily worked younger rocks, such as the oolitic limestones, so important in England, are absent. Indeed, the youngest sedimentary rock types utilised in the building fabrics of the earliest Irish churches, almost without exception, relate to the oldest sedimentary rocks incorporated into the equivalent churches in England and Scotland. With Scotland possessing far more of the older and more difficult to work rock types than England, it is to be expected that the early Irish church more closely resembles the earlier Scottish, than the first English stone churches. In Ireland, therefore, early ecclesiastical buildings exhibit only limited distinctive elaboration or embellishment, and architecturally the evidence which might have helped to designate any changing styles with date is fairly restricted. Attempts to determine an evolution in styles, commenced with some of the earliest authors (e.g. Petrie, 1845, 184; Champneys, 1910, 34) and for these, the variations in the wall masonry of both churches and round towers were noted and examined before being variously interpreted. It should be noted that Leask (1955, 51-52) drew attention to the phenomenon of masonry with ‘vertically bedded’ stones, in churches which he believed, might ‘be assigned to the ninth or tenth centuries’. The subsequent attempts by more modern authors to distinguish different building periods have been briefly outlined in section 4.1. A geological map of Ireland (Figure 4.1) reveals that the rocks at the surface over much of the centre and west of the country are apparently Carboniferous in age, with the major part being described as Carboniferous Limestone.

Berger (1992; 1995) appreciated the possibility of dating those churches with mortared walls by means of radioactive dating. This relied on the radiocarbon-dating of the charcoal preserved in the mortar (the full process being described in Berger, 1995). Some of Berger’s results have been regarded as untrustworthy, but in the present author’s opinion this might be due to the lack of clarity in the sampling techniques (see section 7.1.2).

‘The Carboniferous is an extremely important system in Ireland because it occurs at the surface or beneath Quaternary deposits over nearly half the land area of the country.’ Sevastopulo (1981, 147) The historic term ‘Carboniferous Limestone’ for the most prolific part of the Irish succession of Carboniferous rocks is unfortunate. It represents in Ireland the name of a longish period of deposition of rocks rather than the name for a specific type of sedimentary rock – limestone. Much of the Carboniferous Limestone is, therefore, not actually limestone. This is particularly true for the lower and middle parts of the succession where the rocks are frequently sandy (arenaceous) or muddy (argillaceous). The Carboniferous Limestone includes many different rock types all of which have distinctive properties with regard to their hardness, form of weathering, durability, and manner in which they break when worked for building purposes. A brief summary of these various rock types is given in Table 4.1. In this Table the rocks beneath the Carboniferous are also shown, for the Carboniferous together with the Devonian deposits

With this background of limited confusion concerning the ages of many of Ireland’s earliest churches, the author has now applied the techniques of ecclesiastical geology, and in particular stone bedding orientation and identification, to the problem. It will be discovered that the ‘Patterned’ style, and many of its English and Scottish characteristics, certainly prove to be present in these churches. 4.2 Stone availability in Ireland ‘The geology of Ireland has many affinities with that of highland Britain …There are many facets of the English scene … which are missing in Ireland, partly due to the dearth of rocks younger than the Carboniferous.’ Whittow (1974, 18-19) 79

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.1 A photographic copy of the 1:500,000 scale, Bedrock Geological Map of Ireland, produced by the Geological Survey of Ireland (2006) and reproduced by kind permission. The key to the different geological formations is too complex to display here, but see Table 4.1.

80

CHAPTER FOUR Table 4.1 A summary of the principal Devonian and Carboniferous rocks of Ireland. It should be noted that some of the deposits shown are only regional in occurrence. Information partly from the Geological Survey of Ireland Upper Carboniferous – Pennsylvanian Stage Unit* Depositional Conditions

Westphalian Namurian

72 71

Type of Deposits

Fluvio-deltaic & shallow marine Fluvio-deltaic & basinal marine

Lower Carboniferous – Mississippian 69-70 Fluvio-deltaic & minor marine 68 Marginal marine 67 Shallow water marine Visean 66 Marginal marine 65 Marine basinal 64 Marine shelf 63 Shallow marine & coastal plain

Tournaisian

62 61 60 59 58 57

Reef mudbank Marine shelf Shallow and marginal marine Shallow marine & coastal plain Shallow marine Marine

Shale, sandstone, siltstone with coal Shale, sandstone, siltstone & coal

Sandstone, mudstone, thin limestone Mudstone, sandstone, evaporite Limestone, limestone conglomerate Sandstone, mudstone, evaporite Dark-grey muddy & cherty limestone Limestone & calcareous shale Sandstone, mudstone & conglomerate Pale-grey massive limestone Dark-grey muddy shelly limestone Dark-grey limestone, sandy-mudstone Sandstone, mudstone & conglomerate Shale, sandstone & thin limestone Mudstone, sandstone & thin limestone

Igneous basalt and similar lavas and tuffs Upper Devonian -

55 54

Shallow marine Continental red beds

Middle Devonian -

53

Continental red beds

Sandstone, siltstone & mudstone

Lower Devonian -

52

Continental red beds

Sandstone, siltstone & mudstone

Sandstone & mudstone Sandstone, conglomerates & siltstone

Note * ‘Unit’ of rock type may be related to the 1: 500,000 scale map Bedrock Geological Map of Ireland 2006 (Figure 4.1) for more detail.

provide the vast majority of the building stones in the earlier churches of Ireland.

Lower Palaeozoic, Ordovician and Silurian rocks are present especially in areas on the eastern side of Ireland. Generally these rocks proved unsuitable for early building purposes. Exceptions occur in instances where the finer grained rocks have been altered by levels of metamorphism to slates, phyllites or schists.

The second most extensive suite of rocks in Ireland is the Devonian, typically referred to as the Old Red Sandstone in contradistinction to the New Red Sandstone of more recent Permo-Triassic age which is confined to a small area of north-east Ulster. The Devonian deposits are frequently red, green or purple sandstones and siltstones, occurring in the south and west of Ireland, and in the Dingle peninsula exceeding 2000m. in thickness. Distinguishing fossils are rare in these rocks and it is frequently suggested that the lower parts of the Devonian succession may be of earlier Silurian age.

Although igneous rocks occur fairly widely, only where they reveal clear levels of lineation can their orientation within masonry be determined. The orientation of the basalt lavas of Antrim, for example, proved too difficult to determine within the building material of walls. In the absence of the quality Mesozoic limestones, such as the oolites, and other rock types that can be readily worked; in Ireland, the use of the much harder, more 81

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND intractable, Palaeozoic sedimentary rock types is necessary. It should be noted that when less fissile rocks of comparable hardness, such as silcretes (Potter, 1998), quartzites (Potter, 2002), or flints were employed by the Anglo-Saxon masons of England, they were left in their unaltered shape, or at best, simply broken. Churches in England constructed in these materials preserve few architectural features of the period, but an evolution in improving technique and style of use of each of these individual rock types can be demonstrated.

4.4 ‘Patterned’ workmanship in Ireland 4.4.1 Introducing the varieties Although, in Ireland, the affinity for vertically orientated stones in the early churches remained and can be seen, the range of forms of ‘Patterned’ technique proved somewhat restricted. The use of vertically orientated lineations in stones was much as had been recorded in Scotland, that is, there were instances of vertical orientation in early quoins and arch jambs, and where the rock lithology was appropriate, face and edge bedded stones provided a level of ornamentation. It was possible to use the same abbreviated terminology as had been used in Scotland and England to describe each of these techniques. An important additional feature was evident in the ‘Patterned’ churches of Ireland, this was the inclusion of vertically orientated stones in the antae of the earliest churches (see section 4.4.2).

This very brief summary of Ireland’s geology is given here because, as in England, unlike rock types result in variations in building styles. This is, of course, particularly true in instances of early church building where there was such a reliance on the use of local stone. 4.3 Aspects of ‘Patterned’ workmanship in Ireland Irrespective of rock lithology, but subject to the rock involved possessing some form of distinctive fabric lineation, it is evident that the earliest church builders in either England or Scotland, always attempted to vertically orientate this lineation in elements of their church structures. This ‘Patterned’ style differed according to the rock types of each region. Where the quality Middle Jurassic building stones were available in central England, wall quoins, arch jambs and pilaster-strips were produced to this style. Where the Palaeozoic rocks of northern England and Scotland prevailed, the pilasters were no longer created. In this same area, however, face and edge bedded stones ornamented some of the wall faces; whereas in south and east England ornamentation was provided by coloured bands of rubble stone. What then of Ireland, which most closely resembled Scotland in its rock types?

As at Egilsay Church in Orkney, Scotland (see section 3.3.6), a considerable number of early churches, or more typically their ruins, in Ireland, contain quantities of megalithic, face bedded stones in their wall fabric. Leask (1955) assigned all such churches to the early, that is, ‘Patterned’ period. As the majority of Irish authors would probably disagree with Leask in this view, in this work, each of the churches discussed will be examined according to the evidence available, in an attempt to determine the level of acceptance that should be afforded to the Leask opinion.

In the summer of 2005 the author paid a short visit to Ireland in response to this question. The purposeful orientation of stone bedding in some early churches certainly existed. A further two full summer field seasons followed and the results of these studies now follow. As in England and Scotland, reading the aspects of bedding orientation in wall stones was again burdened with difficulties. Inaccessibility due to height, together with external grime and lichen cover and, internally, lack of illumination and surface washes and plasters, all frequently hid the rock detail. In the absence of freshly broken rock surfaces, wall rock identification proved similarly complex. Detailed scrutiny of the Irish round towers, obviously ecclesiastically related from the proximity to their associated churches, was normally far too complex, for the height, even of their principal common component, the door, was generally too high to examine.

Figure 4.2 A nomenclature for the possible dispositions of stones which might be placed in an anta.

82

CHAPTER FOUR

Figure 4.3 The larger, and more modern, ruined church at Coole, Co. Cork (W 860 952) is viewed here from the southwest. The church, largely built of Upper Devonian sandstones, originally consisted of a single chamber, which today forms the ruined nave, to this the chancel was probably added in the 13th century. The present nave originally possessed antae at each corner, of these only that at the north-west corner now clearly remains. The stones of the remaining anta, with one exception, are all laid BH and are thought to be of post-‘Patterned ‘ style in date. At some date subsequent to the loss of the south-west anta, the quoin (and elsewhere) was rebuilt, this in a much later but similar style to that practiced in the ‘Patterned’ period. 4.4.2 A new nomenclature for the stones in antae

which typically are of a contrasting lithology or different colour from the stones of the wall. Particularly in instances of the casual rebuilding of the quoins of ruined ecclesiastical buildings during this period, sidealternately placed stones were on occasions inserted with their bedding vertical to resemble the ‘Patterned’ style. Such emplacement of this nature can be seen, for example, in the south-west quoin of the larger (or north) church at Coole, Co. Cork (W 860 952) (Figure 4.3), which has been built of re-used stones to replace a previous anta. In Fermanagh, the barrel-roofed mausoleum attached to the church at Aghalurcher (H 366 314) similarly displays in part this late feature in both its doorway jambs and its quoins. Fortunately, in most cases within an individual building, the time interval between the early building style and the later structures which may be similar to that style is sufficient for the two to be distinguishable. There are, however, three instances in the following church stonework descriptions (see sections 4.5.9, 4.5.15 and 4.5. 33) where, without much more extensive examination, doubts remain.

Although the quoins to antae can for simplicity carry the same nomenclature as wall quoins, a single stone, or supplementary stones are often present between the quoins of an individual anta, these stones may be BH, BVFB (bedding vertical face bedded), or BVEB (bedding vertical edge bedded), in much the same fashion as for pilaster-strips. Figure 4.2 illustrates the possible disposition of rocks that may be set into an anta. In this figure the stones are set as they are considered by the author to have originally been placed during the ‘Patterned’ period. In contrast, other antae may have all stones placed BH. Such antae are believed to have been constructed since the ‘Patterned’ period. For a full discussion regarding the possible origin of antae, see Chapter 6. 4.4.3 Scottish practices repeated in Ireland The identification of specific ‘Patterned’ style churches in Ireland proved to present similar difficulties to those found in the churches of Scotland. As in Scotland (see section 3.3.2), post a date of towards the commencement of the 17th century, what might be termed ‘mock-Gothic rustication’ became moderately fashionable (and remains very fashionable in modern Irish domestic architecture). This involves both the framing of windows and the building of wall quoins in side-alternately placed stones

Authors, such as Barry (1987, 139), have drawn attention to the archaeological destruction of Irish ecclesiastical sites caused by the continued use of redundant churches as burial grounds. The problem is not as extensive as in Scotland (Potter, 2006 and see section 3.3.2), but it certainly adds greatly to the difficulties of site interpretation.

83

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND structures reflected the same patterns of stone emplacement as their counterparts in England and Scotland. Initially, geology figured in the considerations, for it was important that, as far as possible, reasonably clear bedding laminations should be preserved in any rocks used in the buildings.

4.5 Examples of ‘Patterned’ style workmanship in early Irish churches 4.5.1. Introduction It is clear that, with the exception of the Romans in the British Isles, early builders in stone relied consistently on very local rock material for construction purposes (Potter, 1999; 2001a; Pearson and Potter, 2002). For this reason, any particular episode of their workmanship in a building tends to be characterised by the presence of a single rock type. This situation differing only where the rock source provided more than one rock type, as it might do from a river bed or deposits of glacial origin.

Nearly 300 sites were examined, the sites being selected according to ease of access, the underlying geology, and to include a range of different ecclesiastical buildings. Only 38 examples of buildings exhibiting critically placed vertically bedded stones are cited in Table 4.2 and located on Figure 4.4. Two or three of these possibly may not represent ‘Patterned’ workmanship. Some 250 plus, therefore, were examined but not listed. Those not cited fall into various categories. The majority showed no evidence of vertically bedded stones in their quoins, arch jambs or antae, if these structures were present. Furthermore, all architectural aspects of these churches indicated a likely ‘Romanesque’ or more recent origin. In

In undertaking this study to determine if ‘Patterned’ style workmanship existed, it was unnecessary to scrutinize the stone bedding orientations in the quoins, antae and arch jambs of every early Irish stone church. The purpose of the analysis was to determine whether Irish ecclesiastical

Table 4.2 An alphabetical list of the ‘Patterned’ churches in Ireland examined in this work LOCALITY Agha Ardagh Ardmore Ardpatrick Banagher Clonmacnoise, T Ciarán Clonmacnoise, T Kelly Drumacoo Dulane Duleek Fore, St Feichin Friar’s Island, Killaloe Glendalough Cathedral Glendalough, St Mary Inisheer, Kilgobnet Inisheer, St Cavan Inishmore, T Benan Inishmore, T Bhreacáin Inishmore, T Chiaráin Inishmore, T MacDuagh Inishmore, T na Neeve Inishmore, T Soorney Kells, St Columba Kilgarvan Killeenemeer Killulta Kilmacduagh Cathedral Kilrush Kiltiernan Labbamolaga (small) Loorha Monaincha Noughaval (small) Oughtmama (west) Ratass Rattoo Temple Cronan Tuamgraney

COUNTY Carlow Longford Waterford Limerick Derry Offaly Offaly Galway Meath Meath Westmeath Clare Wicklow Wicklow Galway Galway Galway Galway Galway Galway Galway Galway Meath Kerry Cork Limerick Galway Limerick Galway Cork North Tipperary North Tipperary Clare Clare Kerry Kerry Clare Clare

GRID REFERENCE S 730 654 N 204 686 X 188 774 R 637 208 C 676 066 N 010 306 N 010 306 M 396 168 N 741 788 O 045 684 N 510 704 R 701 728 T 123 968 T 122 968 L 975 027 L 986 025 L 884 071 L 811 121 L 873 104 L 823 104 L 824 103 L 866 105 N 740 758 W 013 734 R 775 070 R 434 534 M 405 000 R 556 567 M 437 156 R 764 176 M 920 046 S 170 884 R 208 967 M 304 078 Q 853 141 Q 878 336 M 289 000 R 637 830 84

SECTION IN TEXT 4.5.18 4.5.27 4.5.2 4.5.5 4.5.39 4.5.21 4.5.22 4.5.28 4.5.23 4.5.24 4.5.26 4.5.8 4.5.19 4.5.20 4.5.30 4.5.29 4.5.31 4.5.32 4.5.33 4.5.34 4.5.35 4.5.36 4.5.25 4.5.15 4.5.13 4.5.6 4.5.37 4.5.7 4.5.38 4.5.14 4.5.3 4.5.4 4.5.9 4.5.10 4.5.16 4.5.17 4.5.11 4.5.12

CHAPTER FOUR

Figure 4.4 Map of Ireland indicating the positions of the churches described in sections 4.5.2 to 4.5.39. many other churches the decisive stonework was insufficiently clean, or the bedding or lineation traces were too weak (or absent altogether in most igneous rock material), for their orientation to be ascertained.

The church descriptions that follow, although placed in alphabetical order in Table 4.2, will be grouped by region and county below. None of the descriptions is intended to offer either the historical background or the architectural structure of the site or church in detail. They must of necessity be brief, and only deal with and elaborate upon those elements of each church which show evidence of ‘Patterned’ workmanship. In most instances competent descriptions of architectural features for each church already exist.

Omitted from Table 4.2 are a significant number of church sites where a ‘pre-Irish Romanesque age’ has been popularly assumed (such as at the Cathedral or Temple Dowling, Clonmacnoise [N 010316]; Manning, 1998; 2003). Their omission related generally to the state of the stonework; grime or lichen coating making it unreadable, or too few vertical stones to provide reliability; or the stonework lacking precision and probably having been rebuilt.

4.5.2 Ardmore Church, Waterford (X 188 774) Of the four Irish provinces, the province of Munster probably includes the area of the simplest geology. 85

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.5 A simple geological map of Ireland to show its principal geological structures. This map is reproduced from Whittow (1974) by the kind permission of the publishers, Penguin Books.

Extensively exposing just Devonian and Carboniferous rocks, these provide Ireland’s most easily worked building stones, offering a possible reason as to why the province contains evidence of the greatest number (16 – see Sections 4.5.2 to 4.5.17) of ‘Patterned’ early church sites observed. Geologically the region displays more of an east-west (Armorican) fold trend, unlike the rest of Ireland over much of which the rocks trend north-east to south-west reflecting a much older period of mountain building (Figure 4.5).

O’Keeffe, 1992; 2003, 166) describing this church dwell on the excellent Romanesque sculptures that are featured on the exterior of the arcaded west gable wall (Figure 4.6). The lower portions of the walls on the south side of the church are buried, at least in part, by remarkably extensive grave build up (Figure 4.7). For this reason evidence of early wall fabric is best seen on the north side of the church. The distinct change in the style of masonry (Figure 4.8) in the north wall of the chancel was first noted as early as 1864 by Parker. Subsequently, others (such as Brash, 1875; Westropp, 1903; Leask, 1955, 164; 1958, 39-40) have attempted to interpret this fabric change; with Smith (1972) offering the most complete interpretation. The fabric change is visible both on the inside and the exterior of the north wall of the chancel (and rather less well on

Ardmore church is constructed of local Upper Devonian (‘Upper Old Red’) sandstone (Unit 54, Table 4.1); although rarely, as for example in the plinth to the west walls of the nave, Carboniferous grey limestone (probably Unit 61) is incorporated. The majority of authors (such as McNab, 1987; Harbison, 1995; 86

CHAPTER FOUR Figure 4.6 The west gable and the adjoining round tower to the ruined Ardmore Cathedral (X 188 774). The remarkable arcading and ‘Romanesque’ sculpture on the west gable has been extensively referred to in the appropriate literature. The lack of symmetry to the arcading within the wall should be noted.

Figure 4.7 Due principally to grave build up, viewed from the south only the upper portions of the walls of Ardmore Cathedral are evident. Note the string course terminates unusually near the top of the south doorway in the nave wall.

Figure 4.8 Long observed changes in the style of stonework are evident in the north wall of the present chancel of Ardmore Cathedral. Most authors have regarded the larger stones in the lower part of the wall as being of ‘preRomanesque’ age.

the south wall). From the nature of the stonework as seen on the inside (Figure 4.9), this wall has been extended, slightly raised and furnished with a window (now blocked) in a period of rebuilding described by Smith (1972, 7) as Phase IV (‘prior to 1203’). Harbison (1992, 20), provided a date of ‘possibly around A.D. 1000’ for those Irish churches with both nave and chancel, one of his four categories of early churches (Harbison, 1982). He further proposed (1992, 20) that, ‘Under the influence of this innovation, chancels were added to many of the older churches’. In the instance of Ardmore the original small, unicameral church appears to have been

represented by the western portion of the present chancel (and not the nave). At Ardmore, the evidence related to ‘Patterned’ style stone use is restricted and of a ‘secondary’ nature, but it would certainly support the contention that the lower part of the chancel wall was ‘pre-Romanesque’. The stonework of the north-east quoin of the current nave closely resembles in size, shape and character the fabric seen in the lower chancel wall (Figure 4.10), and yet it is set into walls of quite unlike style (see particularly the fabric in the adjoining east wall, Figure 4.11). 87

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.9 The north wall of the present chancel to Ardmore Cathedral viewed from the inside shows different styles of stonework. The blocked window was once clearly inserted into an earlier wall. Note that this wall has also been extended towards the east.

Figure 4.10 Viewed from the north-west, the north-east quoin stones of Ardmore Cathedral nave can be compared with the early lower chancel wall. At the east end of this wall (between the two blocked windows) there is a further fabric change, representing a chancel eastward extension.

Examination of the lower quoin stones reveals that they are, from ground level upwards, orientated; BH (inverted), BVFR (diagonal), BVFR, BH (diagonal), BH*, BH*, BH*, BH*, BH (inverted), BVFL, BH, BH, BH, BH, to the string course which is of ‘Romanesque’ character. These stones have certainly been reset and some of the stones were at that time probably turned to rest horizontally or inverted. The four asterisked stones

Figure 4.11 The north-east quoin of Ardmore Cathedral is here viewed from the north-east. The fabric in the adjoining east and north walls is unlike the stonework in the quoin and the stones were probably, at least in part, re-used from the western quoins of the earlier church.

are probably replacements. The suspicion must be that they originally were part of the quoin construction at the west end of the early church (the present chancel). The better shaped stones which make up quoins are always of premium value, and re-use in this way is common in England.

88

CHAPTER FOUR

Figure 4.12 The north-west anta of Loorha church (M 920 046) is viewed here from the south-west. The anta was in the process of repair and newly pointed. The stone bedding orientations are provided in the text.

Figure 4.13 The lower 1.2m. of the south-west anta at Loorha has been rebuilt; higher, although repaired, the rocks are largely set to the ‘Patterned’ style. course 5, BH, BVFR; course 6, BH, BH; course 7, BVFR, BVFR; course 8, BH, BVFR; etc. In the southwest anta the first 1.2m. above the plinth is rebuilt (Figure 4.13), but above this large blocks of limestone are set in a mixture of BH, BVFR, and BVFL orientations. The newer church has been built into the chancel area and between the eastern antae, but in each, despite repairs, the same mix of orientations can be observed in the face stones of the antae. When visited in June 2009, repairs were being completed by the Office of Public Works and those involved stated that early burials had been uncovered beneath the north-west anta, presumably indicating the presence of a still earlier wooden church beneath the site.

4.5.3 Loorha, North Tipperary (M 920 046) In contrast to Ardmore, the oldest standing church at Loorha (or Lorrha) has received very limited attention. Harbison (1992, 309) suggested that the church is on the site of an early monastery which was founded by St Ruadhán in the 6th century and limited references to the church have been made by Gleeson (1951, 101), Gwynn and Hadcock (1970, 185), Manning (2000) and Ó Carragáin (2005a). Within the walls of the ruined site stands a much more modern Church of Ireland place of worship. The ruins are situated geologically upon rocks of the Carboniferous (Unit 65, Table 4.1), which consist of argillaceous limestones which contain cherts, and these rocks are predominant in the walls.

A barrel vaulted mausoleum has been built into the west end of the nave so that the inside of the west gable is obscured. The south doorway has been described by Harbison (1992, 309) and two infilled doorways occur in the north wall of the nave. In places as much as 2.5m. of the upper nave walls have been rebuilt, as in the north wall, but in the lower walls, areas of original masonry of large limestone blocks are present.

The oldest part of the church is represented by the relatively large nave. An early west doorway has been infilled and noticeably the plinth of simple square section visible beneath the west gable is absent where the doorway was once present. The nave retains its unusually broad (1050-1070mm.) and deep (640-650mm.) antae. Although these have been partially rebuilt, they retain sufficient original, generally megalithic, stones to be considered of ‘Patterned’ origin. The north-west anta stones (Figure 4.12) rise from the plinth (reading from left to right) as: BH; course 2, BH, replaced, BH; BH; course 4, BVFR (with two smaller stones above), BVFR;

4.5.4 Monaincha Abbey, North Tipperary (S 170 884) Monaincha abbey is a moderately complicated church and a number of authors have offered interpretations as to the periods of construction of its different parts (such as 89

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND McNeill and Leask, 1920; Leask, 1955, 129-136; Gwynn and Hadcock, 1970; Cunningham, 1976, 19-30; Harbison, 1992, 310-311; O’Keeffe, 2003, 252). All, it would appear, have been strongly influenced by the phases of excellent ‘Romanesque’ and later workmanship that are evident throughout the building (Figure 4.14). Gwynn and Hadcock (1970, 41) suggested that the church was founded by St Hilary whose death has been recorded as 807. The church site was on an island until the 1790s when the surrounding wetland was drained (Figure 4.15). The local solid geology is of lowest Carboniferous sandstone (Unit 58, Table 4.1, upon which the church is built), with Upper Devonian sandstones in near proximity. The walls of the church appear to have been built of this Carboniferous rock type, although the stones in certain of the ‘Romanesque’ structures are often slightly coarser grained and may include rare small quartz pebbles. On the north side of the church nave, a sacristy with an upper chamber was added about the 15th or early 16th century. The nave and chancel walls sit on a square and almost continuous plinth which below the chancel rises to two (and under the east wall to three) courses. The lowest level of plinth stones has many of the characteristics of certain English, Anglo-Saxon or ‘Patterned’ examples – see for instance, the megalithic stone beneath the northeast chancel quoin. Although the lowest stones in each of the nave quoins tend to be heavily covered with lichen, a ‘Patterned’ style of stone orientation can be determined with difficulty.

Figure 4.14 The ‘Romanesque’ chancel arch at Monaincha Abbey (S 170 884) viewed here from the nave. Note that the relieving arch is not fully continuous and that there is a change in wall fabric style approximately half way up the gable.

South-west quoin: Plinth, BH; stones 1 to 4 inclusive unreadable; 5, BVFR; stone 6 unreadable; 7, BVFR; 8 to 10, unreadable; higher stones BH.

Figure 4.15 The ‘island’ setting to Monaincha Abbey. This view from the north-west shows both the sacristy (left) and the high cross. 90

CHAPTER FOUR

Figure 4.16 The chancel of Monaincha Abbey viewed from the south-east to show the repaired pilasters at the quoins. North-west quoin: Plinth, BVFR; stones 1 to 3 unreadable; 4, BH; 5, BVFL; 6, BVFR; 7 and 8, BH (8 broken); 9, BVFL; 10, BVFR; 11, BH; remainder, unreadable and partly rebuilt. South-east quoin: Plinth, BVFR; 1, BVFL; 2, BH; 3, BVFR; 4, BVFL, 5, BH; 6, BVFR; all higher appear to be BH and replaced. North-east quoin: lowest 1.3m. rebuilt; followed by stones in order, BVFR, BH, BH, BVFR (diagonal), ?, BH, ?BVFL, BH. The two eastern quoins to the chancel are ornamented to rise like pilasters (described as ‘quoin-columns’ by O’Keeffe, 2003, 254). In each, the ornament varies throughout the height and perhaps two or three different periods of insertion are recorded. At the south-east corner (Figure 4.16), the stones rise above the upper course of the plinth as: cushion column base; 1 to 3, all octagonal in outline, cut back on both sides; 4 and 5, round in outline, with cut backs; 6 and 7, as 1 to 3; 8, a round stone, possibly of earlier date than 4 and 5, with a particularly large cut back on the south face; 9 and 10, as 1 to 3; 11, large rectangular capital (replacement). Each stone is apparently set BH. At the north-east corner, the stones rise above the upper course of the plinth, with no cushion but a rectangular base, as: 1 to 5, octagonal and cut back; 6 and 7, round and cut back; 8, octagonal and cut back but not identical to stones 1 to 5; 9, large rectangular capital (replacement). Each stone set BH. The details of these pilaster-like corners are presented here not because they are of ‘Patterned’ style but because their detailed workmanship is of distinct ages. Commencing with the 12th century ‘Romanesque’ threequarter round column, later replacement stones are completed in octagonal outline, a form slightly simpler for the stonemason to cut.

Figure 4.17 The south wall of the chancel at Monaincha and the half-column pilaster near the angle with the nave. The window in the south wall is designed to hold a shutter. built as a half-round column (in which some of the stones are face bedded in ‘Romanesque’ style); presumably to match the earlier columns at the eastern quoins of the chancel. The north chancel wall has been rebuilt where

Near to the south-west angle of the chancel, a further pilaster occurs on the chancel wall (Figure 4.17). This is 91

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.20 Detail of the south jamb of the west doorway to Monaincha Abbey as viewed from the east. Where the outer order of the hoodmould extends into the architrave of the jamb the stones vary, the rounded stones are cut back; those with a rectangular section would appear to be replacements.

Figure 4.18 The ‘Romanesque’ west doorway at Monaincha Abbey viewed from the west.

Figure 4.19 The same west doorway at Monaincha viewed from the east and from inside the nave. There is some visible evidence of repairs or alteration.

Figure 4.21 Further detail of the top of the south jamb of the west doorway at Monaincha. Note the cut back stones. 92

CHAPTER FOUR

Figure 4.22 Ardpatrick church (R 637 208), viewed from the north-west and across the ruined stump of the round tower. The extent of the ivy cover is well illustrated. the corresponding pilaster would have occurred. Internally, the chancel can be observed to have been built against the earlier east nave wall and it seems probable that it even slightly post-dates the elaborate ‘Romanesque’ chancel arch.

In summary, vertically orientated stones built into the four, partially rebuilt, nave quoins may depict the outline of a ‘Patterned’ style church. This church probably possessed a west doorway, now highly modified to a ‘Romanesque’ form. Lower portions of the nave walls show some limited evidence of this earlier church.

The descriptions of others have drawn attention to the interesting windows to the church and also to the ‘Romanesque’ ornamentation to the west doorway in the nave (Figure 4.18). This doorway, with its chevron and other forms of elaborate external decoration (McNeill and Leask, 1920), however, when viewed from the nave, possesses certain ‘Patterned’ style characteristics (Figure 4.19). The south jamb has certainly been altered over time (Figures 4.20 and 4.21), and all the stones comprising the inner face of the jamb are set BH, and more frequently BVFIA, in ‘Patterned’ style. The earliest stones (including the impost and two of the three hoodmoulding stones) are cut in one piece and have a round cross section. The central hoodmoulding stone has been partially repaired and incorrectly orientated in replacement. Cut backs occur on the south side of some of the original stones. The inner face of the south jamb stones carries ornamentation which appears to be of similar ‘Romanesque’ style to that on the outer orders of the door arch. This would seem to entail the unusual in situ carving of the earlier inner order stones. None of these features appear to apply to the north jamb, for all these stones have been repaired more recently and are set BH. O’Keeffe (2003, 254) has proposed that alterations to the west entrance are modern replacements, a conclusion also advocated by Leask (1955, Fig. 73). This is certainly the case for the stones detailed by Leask. Stones in this doorway that have been cut back form a component part of the wall. They are most unlikely to be modern.

4.5.5 Ardpatrick, Limerick (R 637 208) The ruined church of Ardpatrick is situated to the north of the Ballyhoura Mountains, on a distinct vantage point. The church is built of coarse red sandstones, and stands upon rocks of Upper Devonian age (Unit 54, Table 4.1). Exposures of red siltstones near to the church are not used in the church fabric. Only the north and south walls to the west end of the church remain standing to any height. Westropp (1905, 426) and Crawford (1908), both advised that the ends of the church had fallen in 1840. A monastery is claimed to have been founded by St Patrick on the site and Gwynn and Hadcock (1970, 29) also record that a bishop resided at Ardpatrick in 1079. Harbison (1992, 216) stated that the round-headed south doorway, which has now largely collapsed, was ‘built probably around 1200’. To the east of the area that probably marks the position of the south doorway, and projecting to the south, is a small, very dilapidated, barrel roof mausoleum (?) which also possesses a simple roundheaded doorway. The head of a 13th or 14th century doorway, partially created in Lower Carboniferous limestone, is present in the north wall beneath a heavy growth of ivy. This north doorway has been largely infilled and is probably opposite the position of the ruined south entrance (Figure 4.22). Both Leask (1955, 72) and Harbison (1992, 216), as well as others, refer to a western anta. This does not exist. The 93

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.23 The exterior of the south wall of the nave of Ardpatrick church is terminated at its south-west quoin with ‘Patterned’ stone insertion. The stonework viewed here is quite unlike that shown in the north wall (Figures 4.24 and 4.25) and it is considered to be of the earlier ‘Patterned’ period.

Figure 4.24 The stonework typical of the exterior of the north wall of the nave of Ardpatrick church. This wall was probably constructed in the 12th century. Compare with Figure 4.23.

Figure 4.25 The inside of the north wall of the nave of Ardpatrick church exhibits stonework much the same as the outside of the same wall (Figure 4.24), but unlike that of the earlier south wall (Figure 4.23). west gable is almost entirely absent, but a buttress occurs at the north-west and two low buttresses occur at the south-west, corners respectively. The lowest part of the south-west quoin remains visible and its stones are orientated in ‘Patterned’ style. Only five stones can be examined and these read, from the lowest; ?, BVFL, BVFR, BVFL, BVFR.

the south wall (Figure 4.23) is markedly different to that which occurs in the north wall (Figures 4.24 and 4.25). The north wall possesses a slight batter, is about 1.38m. thick and is constructed with well-shaped large blocks of stone supported by numerous spalls. In contrast, the south wall which is presumed to be of the ‘Patterned’ period, is less well coursed, has a greater range of block sizes, which are less well-worked, with few spalls, no batter, and is at least 0.2m. thinner.

Leading east from the south-west quoin, the stonework of 94

CHAPTER FOUR

Figure 4.26 Killulta church, Limerick (R 434 534) viewed from the south-west displays two distinct styles of stonework. That in the south wall (right) may well be largely original, the stonework in the south-west quoin and the west gable being, in comparison, relatively recent. 4.5.6 Killulta, Limerick (R 434 534) The small ruined church at Killulta has been described as the earliest surviving church in the County of Limerick (Harbison 1992). Both Westropp (1905, 386) and Leask (1955, 73) regarded Killulta as being a church of early origin. Unfortunately, much of the west gable collapsed in the 1860s (Quin and Wyndam-Quin, 1865, 270), and has been reconstructed. The fabric variation due to this rebuilding (and elsewhere, such as in the east gable) can fortunately generally be distinguished (Figure 4.26). The original church walls are built of local, fairly wellcoursed, massive blocks of Lower Carboniferous limestone (Unit 62, Table 4.1; some of the rebuilding stone is probably from Unit 61). The square plinth, which has early characteristics, can be traced beneath the walls of all but the collapsed west gable. Of the four major quoins, only that at the northwest corner appears to be largely original. The stone insertion in its lower part exhibits a ‘Patterned’ style rising from the plinth as follows: BVFL, BH, BH, BVFL, ?BH, BVFL, ?BVFR, BH, BH, BH. The top three stones are believed to be replacements. It is possible that the eastern quoins are similarly constructed but the stonework orientations proved too difficult to read. There is some evidence that on the low exterior of certain walls (as the south wall) the stonework courses are in part face bedded. 4.5.7 Kilrush, Limerick (R 556 567) Figure 4.27 Detail of the north-east quoin of Kilrush church, Limerick (R 556 567). The lowest five stones in this quoin above the plinth are set BVFR, BH, BVFR, BVFL, BH in ‘Patterned’ style.

The original Kilrush church, briefly described by Westropp (1900-1902, 168; 1905, 363), cannot have been much unlike the early Killulta church. Harbison (1992, 95

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.28 The west wall of Kilrush church showing the flat-headed doorway. The south (right) jamb of this doorway has five small replacement stones in its lower portion, but higher through stones are placed BVFIA. 223) stated that the church was first recorded in the literature in 1201. However, he concluded that the church ‘is undoubtedly older’. The church is situated on Unit 64 (Table 4.1) of the Lower Carboniferous limestone and this has provided reasonably large blocks, which with the use of a small number of spalls has been well-coursed. The square plinth to the church is nearly complete and it again has an early appearance. Rather more of the standing walls are of early structure than at Killulta, but although much of the south wall appears original, the wall either side of its small modern window (said to have been inserted in 1900), for instance, shows evidence of rebuilding.

not be examined, the east window, however, appeared to have been constructed in the 12th century. 4.5.8 Teampall Molua {from Friar’s Island}, now at Killaloe, Clare (R 701 728) Up until the time of the Shannon Hydroelectric Scheme in 1929 this bicameral church stood on Friar’s Island, about 1km. from its current situation where it was rebuilt in 1930, ‘as a reasonably accurate copy of the original’ (Leask, 1930). On its original site it was described by authors like Brash (1875), Westropp (1900-1902, 159), Champneys (1910), Macalister (1929), and more especially, Leask (1929; 1930). A wide range of authors have also referred to the church in its rebuilt state (including, Leask, 1955, 28, 31; Gwynn and Hadcock, 1970, 82; Harbison, 1992, 65). According to Brash (1875, 19) in the 19th century the nave was already a ruin and the roof of the chancel was in poor repair. Westropp (19001902, 159) published an early print of 1791 which illustrated that, at that time, most of the nave west gable was still standing.

Each of the four principal quoins has been modified to some extent as have the west doorway and the east window. The north-east quoin probably preserves the best evidence of ‘Patterned’ workmanship. Roughly standing in a side-alternate style, the stones rise above the broken plinth: BVFR, BH, BVFR, BVFL, BH, BH, the remainder rebuilt BH, BVFR, BH, BH, BH. The lowest six stones are illustrated in Figure 4.27 and in some of these stones the orientation is poorly visible to the naked eye. In each of the other three quoins elements of rebuilding were evident. Each included one or two vertically orientated stones and it proved impossible to read others.

The rock types typically exposed in the vicinity of Killaloe are Silurian mudstone and greywacke; the blocks of which the church is built are of a dark, greywacke. Lumps of this stone are commonly included within the walls. Much of the church and all structural areas are, however, of sandstone which was probably obtained from the Upper Devonian. The church according to Brash (p.19) originally stood on a ‘low deposit of sand, gravel and boulders’. The majority of the stonework appears to have been quarried and must have been, therefore, transported to the island, although worn boulders are also present.

In the flat-headed west doorway (Figure 4.28), the two through stones in each jamb are probably original. In the south jamb these are placed BVFIA, in the north they proved currently unreadable. The lowest five courses in the south jamb are apparently replacements. Two other doorways, now both infilled, existed at different times in the north wall. Access to the interior of the church was barred so that certain other aspects of the church could

96

CHAPTER FOUR

Figure 4.29 The west face of St Molua church, originally situated on Friar’s Island, now rebuilt in Killaloe (R 701 728). The west doorway to the nave has an ‘Escomb style’ appearance. Although some authors (as Brash, 1875) believed otherwise, the consensus view is that the chancel was at some stage added to the earlier nave. The nave is built upon a simple square plinth and the remains of each of its four quoins show ‘Patterned’ characteristics. The details being: North-west quoin; BVFR, BVFL, BVFR, BVFL. South-west quoin; ?BH, BH, BVFR, BVFL. South-east quoin; BVFL, BVFR (diagonal), BVFL, BVFR, BH. North-east quoin; BVFR, ?BVFL, BVFR, BVFL, BH, BVFL, top two stones replaced. Leask (1930, 131) believed that this north-east corner of the nave had been completely rebuilt ‘in the 19th century or earlier’. In each corner all higher stones are missing.

chancel arch must have been rebuilt at the time when the roof was reconstructed (Figure 4.30).

The west flat-headed, tapering doorway (Figure 4.29), which has probably been rebuilt twice (involving repair prior to the Killaloe move), has well structured ‘Patterned’ jambs: South jamb (outside); the five stones rise above the plinth as; BVFIA (Through stone), BH, BVFIA (Through stone), not determined, BVFIA. The top two courses appear to be replacements. North jamb (outside); four stones rising as; BVFIA (Through stone), BH, BH (Through stone), not determined and possibly a replacement. The lower part of the doorway would in England be stated as being in ‘Escomb style’. Neither the east window nor the south doorway to the chancel, currently offers stonework in which the orientations can be easily read. The position of this south doorway did much to persuade Brash (1875) that the current chancel represented the original church. The

Figure 4.30 The inside of the chancel roof to Teampul Molua viewed towards the east window. This roof has been extensively rebuilt.

97

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.31 The small ruined church at Noughaval (R208 967) has been greatly modified and a barrel vaulted roof has been fitted between its gables. The much larger stones occurring at the church corners appear to be of early workmanship and the quoin stones tend to show ‘Patterned’ orientations. The chancel east quoins, however, again reflect a ‘Patterned’ style. The north-east quoin reads; BH, BVFR, BVFL, BH, ?BVFL, BVFR, roof eaves. The south-east quoin; Repaired plinth, BH, BH (stone broken), BVFL, BH, BVFL, BVFR, BVFL, roof eaves.

4.5.9 Noughaval (small church), Clare (R 208 967) The much altered and ruined small church which resides close to an ivy clad much larger ruined, largely 12th century church at Noughaval has received very limited attention. The whole of the smaller church has at some stage been fitted with a barrel vaulted roof and a central south door. The barrel vaulting has been constructed inside the east and west gables. Still more recently a rectangular window has been placed in the east wall.

Leask (1930), in observing the building being dismantled, but without presenting his evidence, believed that both of these quoins were rebuilt ‘in the 19th century or earlier’. As he could presumably examine the various wall mortars, his observation must be assumed to be correct. Hoey and Thurlby (2004) would have regarded the vaulted chancel roof as being ‘Romanesque’ in style. The chancel walls are considerably thicker than those of the nave. Although the two components of the church (nave and chancel) could have been constructed to the ‘Patterned’ style but be of different ages, this would seem unlikely. It must be assumed that rebuilding of these eastern quoins may have been executed between the 17th and 18th centuries in a relatively recent repetition of the style (see section 4.4.3).

The limestone of which the building is constructed is probably of local origin and from Unit 64 (Table 4.1) of the Lower Carboniferous, but at least two varieties of distinctly different styles of bedding are involved. (Figure 4.31). Although little original stone fabric remains, the lowest stones in each of the principal quoins exhibit ‘Patterned’ style orientations. The north-east quoin; rises to its full height with stones set; ?, BVFL, BVFL, ?, BVFL, BVFR, BH, BH, ?, BVFL, BH. The first six (of seven) stones remaining in the south-east quoin are emplaced above a plinth to include both BVFL and BVFR stones and two (of the four) stones in the south-west quoin are set with vertically orientated bedding (BVFL, BVFR). None of the orientations in the six remaining, side-alternately placed stones in the northwest quoin, could be read with certainty.

At the time at which the Friar’s Island church was being dismantled, Leask (1930, 130) stated that evidence of external wall plaster was observed ‘being found adhering, reversed, to the concealed surfaces of the later abutting walls’. This plaster must, therefore, be adjudged of prechancel age, although it does not seem to have to have been applied during the ‘Patterned’ period.

98

CHAPTER FOUR

Figure 4.32 Artistic licence replaces any real accuracy in this painted wash of the Oughtmama site by Cooke (18423). It shows the presence of the three churches, with presumably the west church to the left (west). The chancel is apparently only represented by one low wall and the alignment of each church has been altered. (Print kindly reproduced from a copy held by Clare County Library). This small neglected ruin is cited here because it illustrates typical difficulties of interpretation. Certain aspects of the ruin such as the east window and the inserted barrel vaulting suggest a relatively modern origin. Only Westropp (1900-1902, 133; 1909) appears to have referred to the building. He recorded the structure (1900-1902) as ‘O’Davoren’s Chapel’ and stated (1909) that ‘a tablet remained in 1839 which we failed to find:“This chapel was built by James Davoren, of Lisdoonvarna, who died 31st July, 1725, aged 59 years”.’ It is now again in the church. He supplemented this information with the comment that it was hard to believe that the building was not older. Although the quoin stones might represent a 1725 copy of ‘Patterned’ workmanship (see section 4.4.3), wall masonry styles differ, suggesting the whole building is unlikely to be of the same date. This creates the possibility that Davoren modified an early ruined oratory to create a personal mausoleum.

chancel, is considered here. At one time it was the parish church (Cooke, 1842-3), and these ruins occur at the westernmost end of the three buildings (Figure 4.32). The churches are situated on, and built of, Lower Carboniferous limestone (Unit 64, Table 4.1). References to the Oughtmama churches include the works of Brash (1875,16), Westropp (1895b; 1900-1902, 130), and Champneys (1910, 105). The nave of Oughtmama (west) church clearly pre-dates the attached chancel and this is reflected in the difference in the wall fabrics (Figure 4.33). The nave itself, however, shows evidence of considerable rebuilding. Of the four nave quoins only those on the north side of the church clearly display ‘Patterned’ workmanship, and this only in the lower part of the wall: In the north-west quoin the stones occur from the base as follows; Square plinth of uncertain orientation, BVFR, BVFL, BVFR, BVFL, BH, ?(possibly re-used), ?, higher stones BH and probably replaced, terminating with a west-facing corbel. In the north-east quoin the succession is similar; BVFR, BVFL, BVFR, BVFL, BVFR, ?, four higher stones BH and replaced, terminating with an east-facing corbel. The lower stones are large in size and placed in side-alternate style.

4.5.10 Oughtmama (Ucht Mama) (west church), Clare (M 304 078) The site at Oughtmama includes a line of three ruined churches which are thought to represent part of a monastic site founded by ‘The Three Colmans’ (Gwynn and Hadcock, 1970, 400; Harbison, 1992, 67). The largest of the three churches, which consists of a nave and

99

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.33 Oughtmama west church (M 304 078) viewed from the south-east reveals distinct masonry fabric differences between the chancel (right) and the largest area of the visible nave wall.

The west splayed doorway is built in a fashion closely resembling ‘Escomb style’. The stones in each jamb rising; BH, BVFIA, BH, BVFIA, BH, the flat-headed doorway surmounted with a megalithic lintel stone (Figure 4.34). Much of the higher west gable wall has been replaced. The two windows set into the south wall are of ‘Romanesque’ style. They were originally identical and externally they are cut to take shutters (Figure 4.35). The chancel arch was probably built into the east wall at the same time as the windows, for the fabric change exhibited in the east nave gable is not unlike that surrounding these windows. This was also the view of Leask (1955, 82), who believed the arch to be no earlier than c. 1150. Although others (such as Westropp, 19001902, 130; Harbison, 1992, 68) believed nave and chancel to be contemporary, the chancel north and south walls abut against the east gable confirming their later date. The eastern chancel quoins, like those on the south side of the nave appear to have been at least in part replaced. The included early stones possess a BH orientation. Brief reference should be made to the two other small ruined churches at Oughtmama. Both show clear ‘Romanesque’ features but these appear to have been imposed on an earlier ‘Patterned’ age building, of which only limited evidence remains. The provision of multiple churches being built close together in a straight west to east line is well recognised in Anglo-Saxon churches in England (Taylor and Taylor, 1965, 14; and see O’Keeffe,

Figure 4.34 The west door to the nave at Oughtmama west church (viewed here from the north-west) possesses jambs in typical ‘Patterned’ orientation.

100

CHAPTER FOUR 1998, 118) (Figure 4.36). It occurs at a number of sites such as those at St Augustine, Canterbury, Repton (SK 303 272) and Jarrow. 4.5.11 Temple Cronan (Teampall Chronain), Clare (M 289 000) This small ruined church in the Burren is dedicated to St Cronán. It has a number of interesting features which have attracted considerable attention. Important amongst these are the ‘Romanesque’ carved heads on the external wall surfaces and corbels (Harbison, 1972a; O’Keeffe, 2003, 202), and the early tomb-reliquaries close-by the church (Figure 4.37). The church has been referred to by a range of authors (such as, Petrie, 1845, 184; Dunraven, 1875, 105; Westropp, 1900-1902, 134; Champneys, 1910, 103; Leask, 1955, 72, 74; Harbison, 1992, 70; Ruffino, 2000). A full account of the ‘Romanesque’ features is provided in The Corpus of Romanesque sculpture (www.crsbi.ac.uk). The church is built of, locally obtained, Lower Carboniferous limestone (Unit 64, Table 4.1). The Corpus description of the church states, ‘The lower walls contain large limestone blocks’. Viewed from the western angles, convincing variations in the fabric of the masonry can be observed and the stones of the immediate corners are considerably larger (Figure 4.38). Furthermore, the lower north-west quoin stones are placed in a ‘Patterned’ style reading from the ground upwards: BVFL, BVFR, BVFL, BH, BH, followed by three ?BH stones and then the west facing corbel. The top two stones and the corbel stone are of slightly different lithology (and presumably of later insertion). The other three quoins to the church have their stones set BH

Figure 4.35 The exterior of the west window in the south wall of Oughtmama west church has from the clear evidence of masonry disturbance been set into an earlier wall fabric. The window has distinct ‘Romanesque’ characteristics and the fitting for a shutter.

Figure 4.36 The close proximity and alignment of the churches at Oughtmama is illustrated in this view, taken from the south-east, of the central and, more distant, west church. 101

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

(exceptions being north-east quoin, stone 2, BVFR and south-east quoin, stone 5, BVFL). Various stones on three of the angles show evidence of ‘Romanesque’ ornamentation (north-west quoin, stone 1; south-east quoin, stones 4 and 5; south-west quoin, stones 1, 2 and 5). The original west doorway is blocked; all authors agree, this probably occurring at the time of the insertion of the ?15th century, north doorway. The jambs to this west doorway have their stones inserted in ‘Patterned’ style. The north jamb stones have their bedding orientated as follows: stone 1, BVFIA (inside) and BH (outside); stone 2, BVFIA (probable through stone); stone 3, BVFIA (probable through stone); stone 4, BVFIA (but probably not a through stone, external stone carved ‘Romanesque insertion and the two of unlike stone sizes); stone 5, BH (thin through stone). The south jamb stones are as follows: stone 1, BVFIA (possible through stone); stone 2, BH (probable through stone); stones 3 and 4 inside, ?BH, but outside represented by one stone of uncertain orientation; stones 5 and 6 inside replacements, outside represented by top jamb stone BVFIA. The doorway is capped with a flat, BH, megalithic through stone (Figures 4.39 and 4.40). Brief reference should be made to the east window, which most authors recognise as extensively altered. That ‘Romanesque’ workmanship is present may be determined from the pellet ornamentation on the large, internal lower jamb stones (stones which may well have been initially of pre-‘Romanesque’ origin, for they are set BVFIA). Externally the window, with just two jamb

Figure 4.37 The small church of Temple Cronan (M 289 000) seen from the south-east. A house (or slab) tomb can be observed in the foreground.

Figure 4.38 Temple Cronan church is here viewed from the north-west and this corner is of larger stones. The blocked west doorway and its replacement north doorway are also visible. 102

CHAPTER FOUR

Figure 4.39 The south-west quoin, the top of the west doorway and carved ‘Romanesque’ heads are all visible in this view of Temple Cronan church. stones each side of a broken single arch stone, have these orientated BH, BVFIA, so that it closely resembles many simple Anglo-Saxon windows in England. The present author would suggest that the window is of much altered ‘Patterned’ origin, with both ‘Romanesque’ and later alterations. 4.5.12 St Cronan, Tuamgraney, Clare (R 637 830) Harbison (1992, 70) describes Tuamgraney as ‘…the oldest Irish church still in use … it is said to have been rebuilt by the Abbot Cormac O Killeen in 969 … and been repaired again … about 1000’. Others describing aspects of the church include Westropp (1900-1902, 154); Champneys (1910, 37); (Leask (1955, 69), Gwynn and Hadcock (1970, 46), Hare and Hamlin (1986), Edwards (1996), O’Reilly (1997) and O’Keeffe (1998, 120; 2003, 66). Portions of the church are ‘Romanesque’ and other modifications are more recent. The chancel of the church is generally described as being of 12th century construction (Leask, 1955, 69) and this is reflected in an abrupt change in the fabric of the masonry (Figure 4.41). The church is built in an area of Upper Devonian dull red and grey sandstones (Unit 54, Table 4.1) which are used in its construction. The flat-headed west splayed doorway to the church can be related to the early ‘Patterned’ origin (Figure 4.42). Seven large stones form the frame, all are cut back to form a surrounding 240mm. wide architrave. As with Anglo-Saxon churches in England the width of the cut back is determined by the size of the narrowest part of one of the stones (Potter, 2006c). All the stones are through stones and each jamb is constructed of three stones: those in the north jamb having bedding

Figure 4.40 The interior view of the west doorway of Temple Cronan church. Compare with Figures 4.38 and 4.39.

103

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.41 The north wall of Tuamgraney church (R 637 830) viewed from the north-east. Pre-‘Romanesque’ walling creates the wall fabric at the west (farther) end, ‘Romanesque’, possibly 12th century, walling at the east. orientations of; uncertain, BVFIA, BVFIA. Those stones in the south jamb were unfortunately too heavily lichen covered to determine definite bedding orientations but they may all be BVFIA. The antae at the western end of the nave are partially preserved. That at the north-west corner is completely rebuilt above the fifth course, with the early stones above the plinth probably laid: BH (single stone); BH oblique (single stone); BVFB (single stone); three stones, orientation of each uncertain; BVFB (single stone). The first four courses of the anta at the south-west corner also preserve their ‘Patterned’ origin. They are set, BVFB (single stone), BVFB (single stone), three stones - ?,?, BH, BVFB (single stone). Stones above the fourth course are a reconstructed jumble. 4.5.13 Killeenemeer, Cork (R 775 070) Like Tuamgraney, Killeenemeer church has been extended towards the east, so that antae remain only at the west end of this ruined church (Figure 4.43). The ruin appears to have been somewhat overlooked but reference has been made to its structure by a number of authors (such as Leask, 1955, 75; O’Keeffe 1998). Ó Carragáin (2005a) states that the church has been ‘rebuilt’, a remark that certainly applies to much of the building. The ruins are probably geologically situated on Lower Carboniferous limestones (as Unit 64, Table 4.1) but the majority of the building stone is of Upper Devonian sandstones and pebbly sandstones (Unit 54), the source of which is not distant.

Figure 4.42 The ‘Patterned’ style west doorway at Tuamgraney church viewed from the north-west.

104

CHAPTER FOUR

Figure 4.43 Killeenemeer church (R 775 070) is viewed here from the south-west to display the west antae. The west doorway, above the plinth, preserves only the lowest two of its original jamb stones on the south side (in the first, with bedding orientated BVFIA, and the higher stone impossible to read). On the north jamb this is reduced to solely the lowest stone (orientated BVFIA). That these stones represent the foot of a ‘Patterned’ style doorway is supported by the evidence of an architrave cut into these three remaining stones, each of which is cut back. Higher stones in each jamb are fairly recent replacements. That the church was influenced by a ‘Romanesque’ period of building (where stones are laid BH) is illustrated by the east window (although this window has probably been rebuilt in more recent times). The eastern extension clearly post-dates the western end of the church. A break in the fabric, best observed in the south wall, both in the plinth, and the lowest courses of stonework above the plinth, marks the beginning of the eastern extension. To the west of this break the lowest two or three courses in both the lateral walls tend to preserve ‘Patterned’ ornamentation (Figure 4.44). In the south wall, the position of the original south-east anta, projecting some 300mm. can be observed projecting beyond the fragmentary east nave wall.

Figure 4.44 Detail of the western end of the northern wall of Killeenemeer church. This portion of the wall is probably of ‘Patterned’ style. Of the three full courses shown the lower two are bedded in normal BH orientation. The third course has the stones inserted with the bedding placed BVFB. The two western antae which extend beyond the west gable 260-270mm. preserve in their lower portions evidence of ‘Patterned’ workmanship. The south-west anta provides the following detail above the plinth (where more than one stone, the record reads from north to south): BVFB; BVFR, BH (2 stones); BH; ?, BVEB, BVFL (3 stones); BVFB; wall higher rebuilt. The north-west anta displays: Double plinth with upper, of two stones both laid BH; BVFB; BH (inverted), BH (diagonal) (2 stones); BH (inverted); BH (inverted), BH (2 stones); BVFB: wall higher rebuilt.

4.5.14 Labbamolaga (small church), Cork (R 764 176) The smaller church at Labbamolaga (Leaba Molaga, Labbamologga) provides only very limited evidence of its ‘Patterned’ style of workmanship. The church described as small and primitive by Harbison (1992, 82) is situated on a site of a monastery which is thought to have been founded by St Molaga in the 7th century (Gwynn and Hadcock 1970, 396). A number of authors provide descriptions or make reference to the church (such as Lynch, 1909, 35-36; Champneys, 1910, 34-39; Cochrane, 1912, 4; Leask, 1955, 61-62; Herity, 1984, 107; Cleary, 105

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.45 The west face of Labbamolaga small ruined church (R 764 176) showing the west ‘Patterned’ style doorway.

Figure 4.46 The north-west anta to Labbamolaga small church. The folder resting against the structure is 310mm. long. Some of the stones in the anta can be seen in this figure to be emplaced BH, and those in the door jambs BVFIA. Compare with Figure 4.47. 2000). The church site is located near to geological outcrops of both Upper Devonian sandstones and Lower Carboniferous limestones, the former being used in the structure of the church (Unit 54, Table 4.1). Some of the rock used in the church is of a fairly conglomeratic sandstone of massive character.

revealing standing feature is the west doorway (Figure 4.45). This is splayed, and constructed of just two stones in its south jamb and only a single stone in its north jamb. These stones are of megalithic size and are through stones constructed of conglomeratic sandstone. They are capped by a flat-headed large sandstone lintel. Cochrane (1912) provided a sketch of 1879 of the west wall illustrating that at the time the doorway may have possessed an architrave, and Champneys (1910, 38) advised that in 1903 the original lintel stone was broken in two by a

With the walls much reduced in height (the east wall is externally only one course high), the evidence for determining the detail of the church is limited. The most 106

CHAPTER FOUR

Figure 4.47 The south-east anta to Labbamolaga small church extends much further from the gable than those at the west end of the church (Figure 4.46). The folder is 310mm long. falling tree. He illustrated (Plate XIV) the absence of a lintel stone in 1910. The lintel stone is, therefore, a replacement. Fortunately, it is the orientation of the bedding in the jamb stones that reveal if the structure is of a ‘Patterned’ character. This is confirmed by the orientation of the cyclopean jamb stones, both of which are orientated BVFIA. The base stone in the south jamb is set BH.

described as a ruin in 1837. The site is situated geologically on Lower Carboniferous rather muddy and flaggy limestones (Unit 61, Table 4.1) and this forms the principal rock in the church walls. As stated earlier, it is not the intention that this work should provide a detailed analysis of the history or archaeology of any specific church. Sufficient here to state that in its overgrown circumstances, no evidence could be seen of anything of architectural definition of earlier than the late 12th century (the east window in the north wall, and the modified east window {Figure 4.48}). In these windows the stonework is in typical ‘Romanesque’ orientation; BVEIA and BH.

At the church corners are the lower stones of antae, those at the west end projecting only 42 and 45mm. from the gable wall (Figure 4.46), those at the east end, 57 and 64mm. (Figure 4.47) from their gable (see section 6.4). The antae are mainly repaired, with no more than the lowest three or four courses in each showing any resemblance to originality. Both the eastern antae show some evidence of being of ‘Patterned’ period origin. In the south-east corner, above the lowest BH stone, the second course of five stones reads BVFL, three small BH stones, BVFR. These two courses are of conglomeratic sandstone. Stones in the two courses above are orientated BH and are of a different lithology, whilst higher stones have been replaced. In the north-east corner stones in the lowest visible course might also be of ‘Patterned’ origin. The western antae exhibit only BH stones in these low courses, from which it must be inferred that they were constructed in ‘Romanesque’ times.

Three of the quoins provide little evidence. That at the south-east is hidden by ivy and other vegetation and the two at the west end have been rebuilt in relatively recent times. A single stone at the base of the north-west quoin is placed BVFL. The north-east quoin (Figure 4.49), however, is constructed of large, irregular, rough blocks of limestone placed in side-alternate fashion with ‘Patterned’ style bedding orientations as follows: BVFL, BVFR, BVFL, position of stone 4 repaired with three smaller stones of different lithology, BVFL, BVFR, BVFL, BH, gable. At the time of inspection, there was insufficient clarity to determine with assurance whether this evidence represented the remnant of an earlier church or untidy replacements in the period of the 17th to early 19th centuries. The stone, in character, is a fairly massive limestone (much as the ‘Romanesque’ windows). This suggests the latter interpretation to be the most probable. The removal of a piece for microscopic examination might help to identify its source and provide information on the period in which it was worked.

4.5.15 Kilgarvan, Kerry (W 013 734) The ruined and partly ivy clad church at Kilgarvan is said to be dedicated to St Garbhan. Archaeologically the site has been neglected, although a complex range of further buildings are present beside the church. A new church was built in the village in 1815, possibly as a replacement, and the church in this brief study was 107

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND 4.5.16 Ratass, Kerry (Q 853 141) The ruined church at Ratass has been described, or referred to, by a number of authors (including, Petrie, 1845, 166; Anon, 1891, 622; Leask, 1955, 69; Gwynn and Hadcock, 1970, 43; Harbison, 1972b; 1992, 184; Fanning, 1977; O’Keeffe, 1998; 2003, 180). Fanning (1977), having dated the decorated monumental stones found on the site, suggested that the earliest part of the church was built around 1000. The remains of the church show evidence of considerable rebuilding, which Harbison (1992) and others, state was undertaken around 1700. Although the church stands in an area of Lower Carboniferous limestones (Units 62 and 64, Table 4.1) this rock type provides only one of the church building stones. Upper Devonian sandstones (Unit 54) are used in the west doorway and the antae, these probably being transported by boat from exposures to the south of Tralee Bay. The west doorway is slightly splayed, with a megalithic, flat-headed lintel (laid BH and actually inverted). In 1910, the doorway was blocked (Champneys, 1910, Plate XV). It is constructed in dull red coarse sandstone and displays an architrave 28-29mm. wide (Figure 4.50). In typical ‘Patterned’ style it is created by making cut backs in the jamb stones of which there are five stones on either side of the doorway. The stones in the south jamb are, from the base, orientated: BH (top only exposed), BVFIA, BVFIA, BH (inverted), BH; and in the north jamb: BH, BVFIA (the bottom edge of this stone

Figure 4.48 The modified east window at Kilgarvan church (W 013 734) viewed from the east. Steel rule length 325mm.

Figure 4.49 Kilgarvan ruined church viewed from the north-east to show the north-east quoin. 108

CHAPTER FOUR determines the architrave width), BH (inverted), BVFIA, BH (but repaired). The east face (inside) of the doorway has been rebuilt and none of the jamb stones passes through the thickness of the wall. Although both west antae are also built of the same red sandstone as the doorway (except for minor repairs), they have both been rebuilt, for their stones are laid, with rare replacement exceptions, BH. The chancel possesses ‘Romanesque’ features, such as the east window (partially rebuilt [see Harbison, 1972b]) with appropriate moulding and a fitting for a shutter. O’Keeffe (1998, 121) suggests that the chancel was added ‘shortly after 1111’. This extension is also in red sandstone so that a similar date for the rebuilt antae seems likely. The chancel east quoins were apparently rebuilt post the 17th century in side-alternate style. Certain areas of repair throughout the church seem to have been completed relatively recently in Carboniferous limestone (Figures 4.51 and 4.52). 4.5.17 Rattoo, Kerry (Q 878 336) Most authors concur with the Harbison (1992, 184) view that Rattoo church ‘was built probably by the Augustinians in the 15th century, though stones from an earlier church are built into the walls’. The earlier monastic foundation is ascribed to St Lughach. Other authors who have referred to this small ruined church are Gwynn and Hadcock (1970, 43), whilst Byrne (1910; 1911) provided a well documented ecclesiastical history. Geologically the church is close to potential rock source outcrops in Upper Devonian sandstones (Unit 54, Table 4.1) and Lower Carboniferous calcareous sandstones (Unit 58), both of which appear in the church walls. Lower Carboniferous thin muddy limestones and more

Figure 4.50 The west doorway of Ratass church (Q 853 141) is built of Upper Devonian sandstone. The doorway has an architrave which has been created by cutting back the various BVFIA and BH stones which form the jambs. The whole structure is typical of work in a ‘Patterned’ style (rule 325mm. long).

Figure 4.51 The south-west nave corner of Ratass church showing the antae. Note that two of the highest stones in the south-west anta are visibly set BVFL and these are thought to be replacements. Most of the later replacement stones are in grey limestone as may be observed high in the west gable. 109

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.53 The north-east quoin of Rattoo church (Q 878 336) viewed from the north. The lowest stones in the quoin are of calcareous sandstone, noticeably larger, and set in ‘Patterned’ style.

Figure 4.52 The south-east quoin of the ‘Romanesque’ chancel of Ratass church which is principally created in sandstone. pure limestones (probably from Units 61 and 62, respectively) found nearby, also occur less commonly in the walls. Architectural features in the church, such as the windows and the west door, can certainly be attributed to Augustinian fabrication, and it is of interest that they all rely on limestone (Unit 62) for their construction. The external walls and the north-west quoin (particularly) are much altered. Large (occasionally megalithic), calcareous sandstone blocks (Unit 58), however, occur at, or towards the bottom of, three of the quoins and these appear representative of the ‘Patterned’ period church. The south-east quoin is largely covered at the base by later structures but the north-east quoin stones can be read as: uncertain orientation, BVFL, BVFR, BH, BVFR with the higher sandstone blocks being replacements of smaller size (Figures 4.53 and 4.54). In the south-west quoin, above a few fragmentary replacements the blocks of calcareous sandstone rise: BVFR, BH, BVFR, BVFL. In June 2009, the church was under repair by the Office of Public Works and, only with difficulty, could the southwest quoin be reviewed. Four or five stones had been uncovered at the base of this quoin but their bedding orientations could not be examined.

Figure 4.54 Detail of stones 2 and 3 in the north-east quoin, shown in Figure 4.53, of Rattoo church. The bedding orientation in these stones is clearly visible as BVFL, BVFR. Subsequent to work on the site in 2009, these stones may no longer be in the same position in this quoin.

4.5.18 Agha, Carlow (S 730 654) The Irish province of Leinster is geologically far more complex than that of Munster. Significant areas of the 110

CHAPTER FOUR

Figure 4.55 Viewed from the south-east the problematical church at Agha, Carlow (S 730 654) displays its two contrasting wall masonry styles. The well coursed and hammered style of granite boulders at the western end of the building would be considered of ‘Romanesque’ origin in other churches. province are of granite, and Lower Palaeozoic (Cambrian, Ordovician and Silurian) rocks are more prevalent than the more easily worked Upper Palaeozoic (Devonian and Carboniferous) that are so widespread in Munster. Agha church is actually sited on rocks of Lower Carboniferous age (Unit 59, Table 4.1) and spalls of this material (shale, limestone and sandstone) are occasionally used in the walls. But, Agha is close to the edge of the Leinster granite and the church is essentially built of this rock type. The granite has been used in the walls in two quite distinct styles (Figure 4.55).

that in the other walls, it consists of irregular, rectangular, uncoursed, granite blocks, the minor gaps being infilled with chips of granite and flaggy impure limestone, in walls that are slightly battered. In the east gable the granite blocks become well coursed above the east window, and in the south wall externally, this change occurs just to the west of the altered triangular-headed window. The church has recently been described in some detail by Ní Ghrádaigh (2008), whilst previous persons who have referred to the church include Leask (1955, 67) and Gwynn and Hadcock (1970, 372). Leask offered an unsupported statement, presumably on the presence of the western antae, that the church had been extended eastwards by one third. Others have all it appears chosen to support this view. Ní Ghrádaigh (2008, 58) presented a supportive assertion, that, ‘The west end of the church … with deep antae (c. 0.55m) and a trabeate doorway … is clearly pre-Romanesque’. It was Manning (1998, 76) who first proposed that the amount which antae extended from the church gable walls might give an indication of the age of the building. He suggested that the older churches possessed the deepest antae. This argument will be examined more fully in section 6.4.2, but it seems generally unlikely to be true.

The granite incorporated in the walls at Agha is relatively coarse in grain size and yet blocks reveal an element of lineation, which it is necessary to explain. Stillman (1981, 98) advised, ‘Flow structures in the granites parallel the margins’ of the flanks of the intrusions. Simply explained, this implies the cooling of the molten igneous granite rock as it was intruded, and the rock material as it was slowly flowing aligning the newly formed crystals. Early stonemasons chose to utilize this crystal alignment in the same way as sedimentary rock bedding. This is evident in just one quoin, the south-east quoin, of the church with the resulting granite block orientations: stone 1 extends beyond the quoin to form a substantial foundation stone (in common Anglo-Saxon style) and is set BH, with stones above it rising BVFR, BH, BH, BVFR, BVFL, BVFR, BH, BVFR, BH, thin shale levelling piece with stones 11 to 17 all BH replacements. The lower part of this quoin therefore has all the characteristics of ‘Patterned’ style work. The fabric in the wall either side of this quoin is quite unlike

There are churches in Ireland with trabeate west doorways which prove not to be of ‘Patterned’ age (such as Mungret larger church [R 544 537]), although it does seem probable that many have been modified later in their history. The flat-headed west doorway to Agha 111

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.56 Agha ruined church is here viewed from the north-west. The two styles of masonry are well exhibited on the inside of the east gable. The inside of the east window is noticeably of ‘Romanesque’ character.

Figure 4.57 From the north-east the ‘Romanesque’ style of the inner Agha church walls is well displayed. Note that above the modified ‘Patterned’ style reset doorway the interior wall structure is of rubble.

church is capped by a large granite block laid BH. The doorway has been extensively altered but it seems possible that, although the doorway has been reset, the large through stones (two in the south jamb, laid BH and BVFIA; one in the north jamb laid BVFIA) may originate from an earlier ‘Patterned’ period church.

Ní Ghrádaigh (2008, 58) regarded the more irregular stonework at the east end of the church as ‘preRomanesque’ but ‘more problematic’. She based her evidence in part on the styles of the two windows. The bedding orientations in the granite blocks of these windows unfortunately proved impossible to read.

112

CHAPTER FOUR

Figure 4.58 The interior of the south wall of Agha church here exhibits both the ‘Romanesque’ hammered granite boulders and the rubble interior wall infill (now cemented into position). The wall fabric at the western end of the church consists of hammered nearly identically sized and rectangular, granite blocks laid in well-structured courses (Figure 4.56). Both Leask (1955) and Ní Ghrádaigh (2008) described these as ‘field boulders’; it would have taken an enormity of time to find such a large number of like shape and size boulders. This technique of construction in Scotland is restricted only to ‘Romanesque’ churches (as at St Blane, Bute; NS 095 535). Each stone has been carefully hammered to be of equal size. It can also be observed that where both the exterior and the interior of a wall is of this hammered stone model (Figures 4.57 and 4.58), a wall may be infilled between these surfaces with a rubble of granite and sedimentary rock pieces, the technique of rubble infill being very much employed in the ‘Romanesque’ period (see section 2.13.3).

end is the older she does accept that the eastern end is ‘pre-Romanesque’. In contrast, the evidence provided largely by stone bedding orientation is that the older portion of the building is the ‘Patterned’ east end with the west end of early ‘Romanesque’ age. This would have involved resetting the trabeate west doorway, a practice relatively common in Norman England. The doorway has certainly suffered much alteration. 4.5.19 Glendalough Cathedral, Wicklow (T 123 968) The Glendalough monastic site (Figure 4.59) is claimed to have been founded by St Kevin in the 6th century (Harbison, 1992, 342). The cathedral is the largest of the buildings which remains on the site and it has been extensively studied (as by, Petrie, 1845, 250; Champneys, 1910; Cochrane, 1911-12; Leask, 1955, 71; 1977; Harbison, 1982; 1992, 344; Hare and Hamlin, 1986; Manning, 1996; 2002; O’Keeffe, 2003, 85, 241). The last but one of these authors provided a detailed and modern analysis of the masonry of the cathedral, and the conclusions from his study are supplemented and largely confirmed by the following examinations of the geological fabric. The principal construction stones for the building are the Lower to Middle Ordovician cleaved phyllite and schist which occurs locally. Leinster granitic rocks are also proximate to the site and these are used in the ‘Romanesque’ and more modern work.

The fragmentary antae at the west end of the building would be ‘Romanesque’ in age, if, following the contention of the present author, the south-eastern portion of the church pre-dates the rest of the ruin. The absence of antae from the east end of the building then becomes difficult to explain. Ní Ghrádaigh (2008) believed she could identify the presence of early antae in the interior of the south wall, in a pair of stones facing towards the east. These are supposed to be evident in her Plate 3.9, but no presence of either an east-facing (church west end younger) or west-facing (church east end younger) anta could be observed by the present author.

Manning (2002, 19) concluded that the cathedral nave incorporated, in part, stones from an earlier nave and that the antae at the west and east gables had been originally components of this replaced nave. Examination of

The article by Ní Ghrádaigh (2008) is extremely detailed (although it contains several geological errors) and although she supports the popular thesis that the western 113

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.59 A general view of the Glendalough site (T 122 968) viewed from the east. ‘bedding’ orientations of these antae reveals their early ‘Patterned’ character. As the rock employed at Glendalough is a phyllite or schist, this lineation is actually caused by the rock cleavage or schistosity, and not the bedding. Reading from the foot of each anta, the details are as follows (it should be recalled that any particular course may involve one or more stones and is read from left to right facing the anta). North-west anta: BH (stone split horizontally); BVFL, BVFR; BH (shows schistose foliation well); BVFL, BVEB, BVFR*; BVFB; BH, BVFR; BVFB; BVFL, BVFB, BVFR; BVFB; level 10, BVFL, broken replacement pieces, BVFR; followed by 5 thin cantilevered capping stones of later date, all BH. Courses 1 to 5 have been chamfered at their edges, except for level 4 BVFR* stone which might be a later replacement. Courses 1 to 8 are displayed in Figure 4.60. South-west anta: BVFL, BVFR (both diagonal); BH; BVFL, BVFR; BH; Diagonal and well foliated*, BH*; BVFB; BH, BVEB, BVFR; BVFB (but diagonal); BVFL, BVFB, BVFR; BVFB; level 11, BVFL, granite insert, BH; BH; followed by 5 cantilevered, capping stones of later date, all BH. Courses 1 to 5 have been chamfered at their edges. The asterisked stones at level 5 appear to have the remnants of head carvings (base only on the northern stone) on their surface. South-east anta: BH, BH (both top only); BH (large, split horizontally); BVFL, BVFR; BVFB; BVFL, BH; BVFB; BVFL, quartzite lumps as infill, BVFR; BH; levels 9 and 10 much replaced; BVFB; three identical later capping

Figure 4.60 Courses 1 to 8 of the north-west anta at Glendalough Cathedral, in which the phyllitic/schistose stone lineations can be observed (see text) and display a ‘Patterned’ style. 114

CHAPTER FOUR

Figure 4.61 Courses 2 (top only) to 6 of the north-east anta at Glendalough Cathedral in which stone lineations show the ‘Patterned’ style with moderate clarity.

Figure 4.62 The outside of the ‘Patterned’ style west doorway at Glendalough Cathedral (Folder 310mm. long).

stone courses. Courses 1 to 5 again show chamfered edges. North-east anta: (lowest stone not wholly visible from east, but from north face BH); BVEB (originally one stone); BH (broken horizontally); BVFL, BVEB, BVFR; BVFB; BVFL, BVFR; BVFB; BH, granite (replacement) above it BH, BH; BVFB; BVFL, BH above replacement granite, diagonal; single BH cantilevered, replacement capping stone. The first 5 directly visible stones are chamfered on each edge with the exception of the 4th visible stone. Courses 2 (top) to 6 are shown in Figure 4.61.

possesses all the characteristics of the ‘Patterned’ period (Figure 4.62). Each jamb probably originally had a similar structure of four stones and this would from the base have read; BVFIA, BVFIA, BVFIA, BH; each being a through stone. On the outside of the doorway an architrave, with a proud face 295mm. wide, created partly by cutting back the cyclopean stonework of the doorway remains present. Modifications include the part replacement of three of the jamb stones (the second course, south jamb outer replacement stone being inserted in ‘Romanesque’ style BVEIA) and the removal of a substantial part of each stone to provide door fixings on the inside of the jambs (Figure 4.63).

An unusual feature of the antae is that they are thinner in structure than the north and south nave walls. To explain this, Manning (2002) proposed that the antae in the current nave previously served as antae in an earlier, smaller church on the site that was completely dismantled to provide stone by stone replacement antae for the newer nave. Thus the record of ‘Patterned’ workmanship in these features would represent the style of an earlier nave. A date of late ninth or early tenth century has been proposed by Manning for this early stone church. The original west doorway, despite some modification, also appears to have been reused in the new nave for it too

The style and dating of the wall fabric of the present nave have provoked much interest. The lowest three courses are created from well-squared, tightly packed ashlar blocks of phyllite/schist, above which the wall is of moderately coursed rubble masonry (Figure 4.64). The time gap between the two fabric forms can be debated, from two separate building phases (Leask, 1955, 71) to the view held by Manning (2002) that it was negligible. Manning argued his case on the large number of joggled joints in the lower levels (these being related to re-use of the earlier nave stones), the presence of obvious re-used stones in the upper levels, and particularly, the absence of 115

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.63 The inside of the same ‘Patterned’ doorway illustrated in Figure 4.62. The jambs of the frame have been cut into to provide a setting for a door at some post- ‘Patterned’ date.

Figure 4.64 The north-east portion of the north nave wall at Glendalough Cathedral. This illustrates the two masonry styles with the lower three ashlar courses emplaced in ‘Patterned’ style. any of the ashlar type blocks in re-use in the upper levels. The joggled joints argument is weak, for the erection of the lower wall from phyllite/schist, where in many instances the rock is cut across its lineation, would certainly have involved the mason with difficult shapes to adjoin (i.e much ‘joggling’) (Figure 4.65).

same wall order is BVFB, BVFB, and BH (where present). Manning (2002) stated that, ‘On balance a tenthcentury date cannot be ruled out’, and the construction of the replacement nave and this lower masonry workmanship is of obvious ‘Patterned’ style. The difference in the widths of the nave walls to the antae becomes (together with the absence of ashlar blocks in the upper wall levels) the rationale for arguing for the existence of a first and then a second nave. This argument then requires the careful repositioning of each of the first nave stones in the same original orientation when they were placed in re-use in the newer building.

Careful examination of the lowest three courses indicates that in areas such as the nave walls, the stones are sometimes laid to a pattern. On the exterior of the south wall, for instance, the first course is laid BH, the second BVFB, and the third, where present, BH. Internally, the

116

CHAPTER FOUR

Figure 4.65 This joggled joint occurs on the inside of the west wall of Glendalough Cathedral. (Rule marked in centimetres). On analysis, Manning’s arguments (2002) to explain the nave walls and antae thickness differences, involving such careful re-use of stones, seems unlikely. It appears much more probable that all the ‘Patterned’ workmanship (in the antae, low in the north and south nave walls, and in the west doorway) is of the same age. These then define the size of the original ‘Patterned’ church. This simpler and more orthodox explanation is not impossible if the method of construction of the antae and the rationale for their use, as described in Chapter 6, is accepted; for this does not require the nave walls and the antae to be the same thickness.

Chapter 6). The lineation (BH) and shape of the cantilevered stones at the top of each anta suggests they are of later origin than the ‘Patterned’ period of construction (Figure 4.66).

In his 2002 paper, Manning touched on a subject which he failed to follow to a conclusion, when referring to the antae he wrote, ‘… they were widened at the top to the full thickness of the wall, indicating that the builders were conscious of this disparity and wished to have the upper bearing surfaces of the antae as wide as the walls of the church to hold the end rafters of the roof.’ Manning (2002, 20) From this one could infer that Manning was proposing that the purpose of antae might have been to support a short protective porch over each gable. This is contrary to the strongly held view of many, that antae were carried up the gables to beyond the roof ridge-piece as portrayed in the newly reconstructed island church of St MacDara (L 72 30) and reconstructed and partly evident at the church at Kilmalkedar (Q 403 062). Others, following the Manning description, interpret the antae as serving the role of corbels and carrying roof supporting timber up the gables (For a full discussion on the origin of antae, see

Figure 4.66 The BH capping stones at the top of the south-west anta at Glendalough Cathedral. 117

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.67 The ruined church of St Mary, Glendalough (T 122 968) viewed from the north. In this brief analysis a number of other small points, partly of correction, should be made. The chamfering of the lower antae stones is in the present author’s belief a ‘Romanesque’ alteration and was undertaken, as is commonly the practice in England, in this period. The modification was performed on the in situ ‘Patterned’ antae stones. Similar style chamfering can again be observed in the obviously ‘Romanesque’ granite built chancel arch jambs of the Glendalough building. This chancel arch carries a few re-used blocks of phyllite/schist at the foot of its jambs presumably from the earlier nave east wall. The sanctuary to the south of the chancel is probably later in date than the chancel for it is constructed of different types of stone.

described the various exposures that were then visible at Dundry and his ‘Dundry Freestone’ description again resembles the rock visible at Glendalough. A recent visit to the Dundry area confirms the rock identification. 4.5.20 St Mary, Glendalough, Wicklow (T 122 968) Harbison (1992, 344) suggested that the nave of the church of St Mary, Glendalough might be of 10th or 11th century and that a chancel was added to create the present church (Figure 4.67). Various authors, commencing with Petrie (1845, 167) and Champneys (1910, Plate XVI), have made reference to this church. In contrast to the Glendalough Cathedral, the use of phyllite and schist in the walls and church structures appears to have been relatively late, and in general, to be subsequent to the employment of blocks of granite. Only in the nave plinth is phyllite or schist used in an early stage of building. The reason for its use below the granite must surely relate to the potential difficulty of creating suitably shaped plinth stones in granite. Wherever they occur in the church wall structures, pieces of phyllite/schist are placed with a ‘post-Patterned’ orientation of BH.

Manning (2002) noted that the ornamented ‘Romanesque’ arch work in the chancel area consisted of imported Dundry Stone, from near Bristol. This rock was used in England from Roman times. The stone used is a light cream, Middle Jurassic recrystallised limestone and unusually it contains no obvious ooliths, nor any evidence of fossils. Without microscopic examination the exact origin of the stone is difficult to confirm, but the source seems possible, for Waterman (1970) referred to this stone being imported into south-east coastal localities of Ireland over the period c. 1175-1400. To be utilised at Glendalough the stone must have been transported via the River Avonmore in small boats. Waterman (1970, 65) described some of the Dundry Stone used as follows, ‘…the rock has a granular, rather crystalline appearance’. This description is not unlike the material at Glendalough. Nearly a century ago, Reynolds (1912)

In contrast, the more difficult to read, lineation orientations in the granite blocks may exhibit in some instances a vertical attitude. In each of the four quoins to the nave, the rocks used are a mixture of phyllite/schist and granite, almost certainly from their manner of use indicating a structure that has been rebuilt. One or more of the granite blocks in each quoin can be seen to be placed in a vertical orientation. 118

CHAPTER FOUR site as at the Priest’s House, with rare pieces also seen in the cathedral. The rock is presumed to have been quarried in Antrim.

Figure 4.68 The west doorway to the church of St Mary, Glendalough has its stones set in ‘Patterned’ style. The doorway is viewed here from the inside.

Figure 4.69 The outside of the west doorway at St Mary, Glendalough preserves its architrave the width of which has been determined by the north (right) jamb stones.

The west doorway is built of large blocks of granite and shows little evidence of alteration (Figures 4.68 and 4.69). It is splayed, with a flat-head and a large granite lintel, there are just three through stones in each jamb, with foliations set either BH or BVFIA (although it proved impossible to read all stones). The granite blocks on the exterior surface are cut back to create an architrave, which tapers slightly up the door jambs (from 160-170mm. to 145mm.). The architrave continues across the lintel stone with a further cut back, and on the underside of the lintel a cross has been cut (Figure 4.70). All features in the doorway indicate a ‘Patterned’ origin. The chancel and small modifications in the nave reflect workmanship of ‘Romanesque’, and in some instances later, style. It should be recorded that the Middle Jurassic recrystallised limestone from England seen in the cathedral is used sparingly in the ‘Romanesque’ work observed at St Mary. The exterior of the east chancel window and the base stones of the much altered jambs to the north doorway are constructed of this material (Figure 4.71). Repair work executed in the more recent centuries (possibly as recently as 1870) has been completed with sawn blocks of scoriaceous and vesicular basaltic lava. This stone has been used in the south nave window (Figure 4.72) and is found elsewhere on the Glendalough

Figure 4.70 The cross which is carved on the underside of the granite lintel to the west doorway of St Mary, Glendalough.

119

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND 4.5.21 306)

Temple Ciarán, Clonmacnoise, Offaly (N 010

Clonmacnoise (or frequently Clonmacnois) is thought to have developed as a monastic site following its foundation by St Ciarán in 548-9 (Harbison, 1992, 276). The site and its component buildings have been extensively described by many authors; these include Petrie (1845), Westropp (1907), Macalister (1909), Champneys (1910), Leask (1955), Manning (1998; 2003) and Fitzpatrick and O’Brien (1998). The solid geology of the Clonmacnoise site indicates that the area is established on rocks of Lower Carboniferous age (Unit 61, with Unit 62 nearby, Table 4.1). However, the earlier rubble walls of many buildings on the site frequently contain fragments of other rocks in addition to these muddy limestones and limestones. These were apparently collected from fluvio-glacial deposits at or close to the site. Sandstones and quartzitic sandstones originally derived from Upper Devonian deposits, and calcareous sandstones from the Carboniferous (probably Unit 60), are found in various walls. The stone variety seen at Temple Ciarán (or Clonmacnoise Cathedral) is closely comparable to that observed in local gravel pits working the fluvio-glacial gravels (as at N 046 202). Temple Ciarán (Figure 4.73) was described in some detail by Manning (2003). His description of the building as ‘in a very poor state’ is certainly apt. The stone bedding orientations tend to confirm Manning’s findings as to the building’s structure. Stonework in the south-west quoin, for instance, is laid BH confirming its modern character (Figure 4.74). The orientations of the stone bedding in the three antae which remain, suggest that Manning was over-optimistic in his interpretations as to originality. The north-west anta carries a few stones only that have their bedding vertically orientated, the south-east appears to have been completely rebuilt, with only the north-east anta over courses 1 to 7 including sufficient BVFR and BVFL stones to be regarded as original at its base (Figure 4.75). The doorway preserves very little evidence of originality. Initially the body of the building was probably built of fluvio-glacial debris and, therefore, Carboniferous thinly bedded limestones and calcareous sandstones and Devonian sandstones and quartzites made up a mixed fabric. Because of this, early alterations, completed in similar materials are difficult to differentiate.

Figure 4.71 The base of the west jamb of the north doorway to St Mary, Glendalough has been built of the same Middle Jurassic limestone as seen at Glendalough Cathedral. It is possibly Dundry Stone from near Bristol.

Macalister (1909, 141-142) suggested a date of c. 920 for the construction of the Temple, and Manning (2003, 69) proposed a date of c. 909. The specific date suggested by Manning was based upon his view that the walls that are pierced by putlog holes resemble those in the north wall of Clonmacnoise Cathedral which he believed could be accurately given that date (Manning, 1995b). The present author would object to the use of putlog holes for the purpose of correlation of dates, for they are found in walls of buildings of any age up to the modern use of free-standing scaffold. The likeness in putlog style might, however, mean that the two buildings were repaired or repointed about the same time. The geological

Figure 4.72 Replacement stones of basaltic lava are difficult to distinguish from the phyllite/schist in this window in the south wall of the church of St Mary, Glendalough, viewed from the exterior. They make up most of the frame. 120

CHAPTER FOUR

Figure 4.73 Temple Ciarán, Clonmacnoise (N 010 306) viewed from the south side, with the River Shannon and other ecclesiastical buildings in the distance.

Figure 4.74 The west gable and doorway of the ruined Temple Ciarán. Looking ‘old’ it preserves little that is original fabric.

Figure 4.75 The east gable and antae of Temple Ciarán. The north-east anta (right) preserves elements of its ‘Patterned’ originality in its lower stones. Putlog holes are evident in the east wall. 121

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.76 The south doorway of the Dulane church ruin (N 741 788) viewed from the north (interior). Note the wall fabric which is quite unlike that in the west wall (Figure 4.77). The wall in which this doorway occurs has probably been rebuilt. interpretation of the stonework provides sufficient information to confirm a ‘Patterned’ origin. Berger (1995, 169-170) provided a radiocarbon dating on some of the mortar to offer a range of dates (AD 660-980) encompassing these proposals as to the building’s date of foundation.

4.5,23 Dulane, Meath (N 741 788) Dulane church preserves little more than the west gable with its antae. Authors describing these ruins include Dunraven (1875, 94), Champneys (1910, 36, Plate XVII), Leask (1955, 67), Gwynn and Hadcock (1970, 35) and O’Neill (2002, Fig. 11). Gwynne and Hadcock suggested that this church may have flourished as early c. 450 and was burnt on a number of occasions, the first being in 920 (Manning, 2000). The church is set on a mound in a circular graveyard and also shows the influence of considerable grave build-up. Geologically the site is founded on Lower Carboniferous muddy limestones and sandstones (Unit 60, Table 4.1) but Silurian mudstones are not distant from the site.

4.5.22 Temple Kelly, Clonmacnoise, Offaly (N 010 306) Only fragmentary walling of Temple Kelly remains with the quoins standing no more than two courses high. Westropp (1907, 289) advised that the east gable ‘stood in 1738’. The building is normally described as having been constructed to replace a wooden predecessor in 1167 (Westropp, 1907; MacDonald, 2003). Those walls remaining are of the order of 0.8 to 0.85m. thick, with the evidence for an early ‘Patterned’ origin being limited to the quoins. The stone bedding orientations at the three remaining quoins is as follows: north-west quoin, two stones, BH, BVFR; north-east quoin, two stones, BVFL, BVFR; south-east quoin, single stone, BVFR. All the quoin and wall stones are of Lower Carboniferous limestone. This detail is obviously weak for it could, of course, be possible that all these quoin stones have been set in modern times.

The antae and that part of the south church wall that still stands all appear to be largely rebuilt. In the south wall a 13th century or later doorway has been inserted (Figure 4.76). In this, generally limestone, wall, Silurian mudstone fragments are included. The antae and the west wall are constructed almost entirely of limestone (one sandstone block occurs in the northern anta) and all stones but one appear to be laid BH. In the south-west anta the lowest stone would appear to be original, it is face-bedded (BVFB) and of calcareous sandstone.

If, as suggested above, the Cathedral at Clonmacnoise is regarded as being of comparable age to Temple Ciarán (Manning, 2003), examination of the north Cathedral wall would have been desirable. Unfortunately, it has proved impossible to scrutinise the stonework of this building for on the occasion of every visit to the site the Cathedral has been under renovation.

The west doorway is constructed of megalithic blocks of massively bedded sandstone (Figure 4.77). The doorway is flat-headed with a truly cyclopean lintel stone which is now broken. The lower part of the door jambs are covered with earth so that only the top four jamb stones can be examined (the lowest in part). On the south side the three uppermost stone orientations read; BVFIA, 122

CHAPTER FOUR

Figure 4.77 The west wall of Dulane church. The west, like the south doorway, cannot be seen to its full height because soil covers the threshold to some depth. BVFIA, uncertain, all being through stones. On the north jamb the orientations are, uncertain, BH, BH, BH, all but the top stone being through stones. Two small semicircular grooves are cut, at approximately 50 and 85mm. from the inside front edge of both the jambs and the lintel stone to create an ornamental architrave; this being repeated on the front inside face of the jambs and lintel (O’Keeffe, 2003, 309, note 95, presents an alternative view). One of the south jamb stones has been used as a whetstone for knives, from which it is possible to estimate the quantity of infilled soil above the threshold of the doorway. From the information provided by the limited jamb exposure and the general style of the doorway, the work must be accepted as being of late ‘Patterned’ style. The inside face of the doorway has been modified to fit a door.

Lower to Middle Ordovician slates are, however, present nearby. Only three walls of the church stand, the west wall and quoin stones having been removed. The building stands on an inclined site which falls away to west and north (Figure 4.78). As a result the limestone wall on the north side of the church exposes its slate plinth and significant foundations on the wall exterior (Figure 4.79). Bradley (1980-81) thought that these foundations might represent evidence of an earlier church. Of the two quoins that are partially preserved only the stones in the south-east quoin could be interpreted. These are placed in side-alternate pattern and from the base read; BH, BVFR, BVFL, BH. The arched and slightly splayed south door has jambs which, although of small limestone blocks, tend to a long and short style. The west jamb reads; plinth, BVFIA, uncertain, BH, BVFIA, BH. The east jamb has four small BH replacement stones at the base, and then rises, BVFIA, BVFIA, BH, BH. Some of the voussoir blocks in the arch are parallel-sided, typically a feature of ‘Patterned’ workmanship. That the church is old can be related to the manner in which the land now falls away to the north; evidence that it was probably built in the ‘Patterned’ period seems reasonably certain from the limited support provided by the stone bedding orientations.

4.5.24 St Cianan (St Patrick), Duleek, Meath (O 045 684) Authors frequently make reference to the church of St Cianan, Duleek for the name of the village is believed come from the ancient Irish daimhlaig used to describe a stone church (Petrie, 1845, 141). The site has been variously attributed to being that of the first Irish stone church (Hamlin, 1984, 118). Harbison (1992, 258) proposed that this could have been as early as the 5th century although in 1982 (p. 620) he had been more inclined to accept a later date. Bradley (1980-81) recognised this site, which is north-west of St Cianan churchyard, as that ‘more correctly identified with St. Patrick’s Church’. Manning (2000), from the Annals, listed the record of the church of Duleek being destroyed in 881, but it is suggested that this was probably a wooden church. An ecclesiastical history of Duleek is provided by Cuffe (1963; 1965). Geologically the Duleek area is situated on rocks of Lower Carboniferous limestones and calcareous shale (Unit 64, Table 4.1).

4.5.25 St Columba, Kells, Meath (N 740 758) Generally described as St Columba (St Columb, St Columcelle, St Columcille, St Colmcille)’s ‘House’, this building is considered to have been, at least in part, an ancient oratory. The saint established a monastery in Kells at the beginning of the 9th century (Harbison 1992, 261), and the locality is famous for the illuminated Book of Kells (which a few have claimed was completed in Iona). The building has been described by a wide range of 123

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.78 The church of St Patrick, Duleek (O 045 684) is here viewed from the south. There is today a significant fall in land to the west (and north). century; and radiocarbon dates from the lower portions of the fabric by Berger (1995) provided dates in the ranges of 540-1020 and 610-980. The carbon dating led O’Keeffe (1998, 1160) to attempt to reconcile the confusion with the remark that this was ‘…a building with a later superstructure of probable eleventh- or twelfth-century date’. Geologically, the rocks in the region are of Silurian mudstones and greywacke, but not distant, Lower Carboniferous outcrops of limestones and sandstones are present. Examination of the building’s fabric immediately confirms O’Keeffe’s suggestion. Low in most walls, but at variable height, there is a marked change in the style of masonry (Figures 4.80 and 4.81). This is also obvious inside the building (Figure 4.82). At the lower levels irregular lumps of rock and past river cobbles and boulders are set widely apart in ample mortar. This is very much to the style of many Anglo-Saxon buildings in England, where the rock breaks into, or occurs in, lumps that lack parallel sides. The boulders are of both limestone and sandstone from the Carboniferous. The older style wall continues a short distance up the south-west quoin to provide stone bedding orientations for the four lowest stones of; BVFR, BVFL, BH, BVFL. The fifth stone is also BVFR but it is of sandstone and possibly a replacement. Higher stones are all BH (with one exception). In the north-west quoin, where an early square plinth (BH) can be seen, although two BVFL stones are used, the whole quoin is much altered and almost entirely of BH stones. The eastern quoins similarly fail to preserve ‘Patterned’ workmanship. The original west doorway appears to have lost both its jamb and lintel stones and has been blocked.

Figure 4.79 A remnant north-east quoin and the visible north plinth below the north wall at Duleek church authors including Petrie (1845, 171) and Leask (1955, 33). Opinions as to the age of the building have fluctuated widely. Authors (such as Hamlin, 1984, 120; Herity, 1984, 110; Edwards, 1996, 124) have proposed that the building dates from the 12th century; Leask (1955, 33) suggested the year A.D. 814; Laing (1977, 173) early 9th

The studies of Hoey and Thurlby (2004) would suggest that the vaulted interior to the building is unlikely to be 124

CHAPTER FOUR

Figure 4.80 St Columcelle’s ‘House’, Kells (N 740 758) is here viewed from the south-west. The wall fabric change near the base of the south wall can be distinguished. The south door is reasonably modern and now provides the access to the building.

Figure 4.81 A closer view of the south wall of the Kells building reveals more clearly the nature of the older wall fabric which includes river derived cobbles and boulders of Carboniferous limestone and sandstone. much earlier than 12th century. The roof covers an upper and lower floor and a metal ladder can be used to access the interior of the high-pitched roof. Leask (1955, 27-33) discussed this roof and its structure at some length for he used it to represent a stage in the evolution of the Irish stone roof, in the belief that it was of early 9th century date. The ‘propping arches’ described by Leask are

shown in Figure 4.83. O’Keeffe (2003, 90-91) presents the view, fully supported by the present author, that the roof must be of much the same age as the internal vault. Internally, on the inside of each gable there are two clear changes of fabric; the upper of these at the commencement of the vaulting introduces coursed stone of later, probably post 12th century date (Figure 4.84). 125

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.82 The interior of the north wall of the Kells building also shows the same fabric change. The lower stones are spaced with mortar and less well coursed.

Figure 4.83 The ‘propping arches’ which are built into the interior of the roof of St Columcelle, Kells.

Figure 4.84 The interior of the west wall at Kells displays a fabric change at gable level, indicating rebuilding at the time of construction of the vault and roof. 126

CHAPTER FOUR

Figure 4.85 A view of the Fore monastic site (N 510 704) from the south-west, with St Feichin church in the foreground. 4.5.26 St Feichin, Fore, Westmeath (N 510 704) Each of the antae shows evidence in the lowest few courses of originally being constructed to the ‘Patterned’ style. In the north-west anta this is displayed only in the first three courses which rise: BV (direction uncertain); BH (replaced on north edge by sandstone); BVFB. The south-west anta probably contains original stones up to, but not including, the fifth course, but above the single stone in the first course, placed BVFB, the stones have been badly damaged and partially replaced. In the southeast anta, only two original courses exist and in the north-east, three, with various vertically orientated bedded stones in each part of these antae. In the rebuilt portions of the antae, stones are smaller, frequently of rocks like sandstone and calcareous sandstone, and set BH. A number of face-bedded stones are present low in the north nave wall and this probably represents an area of original masonry.

A monastic site (Figure 4.85) is believed to have been established by St Feichin at Fore as early as A.D. 630 (Harbison, 1992, 330). Various authors, including Petrie (1845,171), Champneys (1910, Plate XV), Leask (1955, Fig. 37, 69), Gwynne and Hadcock (1970, 36), Masterton (1998; 2000), Sharkey (1999) and O’Keeffe, (2003, Fig. 17) have made reference to the early church. This is situated on Lower Carboniferous muddy limestones which contain occasional cherts (Unit 65, Table 4.1), and these may be observed in the church walls and the cliffs above the church. The earliest building at St Feichin, Fore is the nave which carries antae. To this a 12th century chancel was added to create the present church. The west doorway is of typical ‘Patterned’ style; flat-headed and splayed and constructed of megalithic stones (Figure 4.86). Each jamb is constructed of three, massively bedded limestone blocks, the bedding orientation of the jamb stones is as follows: north jamb; uncertain (interior, includes small replacement stones), BVFIA (through stone), BVFIA: south jamb; BVFIA (through stone), BVFIA (nearly a through stone, but partly replaced on the inner face by sandstone: externally, the base determining the width of the architrave), BH (through stone), minor replacement stones followed by the lintel. The lintel stone, is orientated BH, and carries a carved Greek cross above the architrave. It has at some time been reset and the wall immediately above the doorway and to the south of it has been rebuilt. The architrave has been created by cutting back the large stones. Early photographs (see Woods, 1907) indicate that most of the repairs to the doorway have been over the last century.

Considerable areas of both the nave and the chancel of St Feichin provide examples of rebuilding of different periods. The eastern quoins of the chancel have, for instance some time post the 17th century, been rebuilt in part in side-alternate style (see section 4.4.3). 4.5.27 St Mel, Ardagh, Longford (N 204 686) St Mel is believed to have been a nephew of St Patrick, and the church at Ardagh was perhaps, therefore, established in the mid-fifth century. Today, the ruins show a significant amount of alteration and possibly for this reason their study has been limited. Harbison (1982, 628) refers to unpublished work by de Paor in uncovering traces of a possible wooden church beneath the present building of stone. The Ardagh site is geologically situated 127

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND on Lower Carboniferous muddy limestones of Unit 65 (Table 4.1); close by, lower deposits in the Carboniferous provide calcareous sandstones and conglomerates (as Unit 59) which can be found also in the church walls. The church possesses antae which have been very much rebuilt. Only the lowest two courses of the western antae, above the prominent plinth, have not been reset. In the north-west anta the largish calcareous sandstone blocks involved are placed; BVFL, BH; and in the second course a single stone, BH (Figure 4.87). Although higher stones are in part placed with their bedding orientated vertically they have been adjusted and in one instance a conglomeratic sandstone has been introduced. The southwest anta is again of calcareous sandstone blocks, laid BH; with second course, BVFL, BVFR, and higher rocks again disturbed, showing joggled jointing and a different mortar (Figure 4.88). The south-east anta (Figure 4.89) is constructed entirely of calcareous sandstone with all but one stone of the lowest course laid BH, and the north-east anta has apparently been almost completely rebuilt with blocks of calcareous sandstone (one stone in the lowest course is set BVFL). Calcareous sandstones comprise the major constituent in the wall fabric, although the upper portions of all walls have been repaired and built up, particularly with pieces of muddy limestone. The lower part of the external face of the east gable wall is completed in megalithic blocks of face-bedded, slightly calcareous sandstone (Figure 4.89) suggesting that this wall may be of a ‘Patterned’ date. The wall plinth which can be seen below all walls appears to normally be of slightly calcareous sandstone, but it has been replaced at the south-west corner with

Figure 4.86 The west doorway of St Feichin church, Fore is constructed in ‘Patterned’ style. Note that there is a change in wall fabric to the south (right) of, and above, the doorway.

Figure 4.87 The west face of St Mel, Ardagh (N 204 686) and the north-west anta. 128

CHAPTER FOUR conglomeratic sandstone. The flat-headed west doorway (Figure 4.87) has jambs of a selection of rock types and little originality. In summary, this church appears to have been built in the ‘Patterned’ period with blocks of calcareous sandstone (of variable lime content) from the Carboniferous. At some subsequent time, conglomeratic sandstones were used for replacement purposes, as in the inner door lintel and the south-west plinth area. More recent replacements of smaller muddy limestone pieces may be seen, as for example, in the inner west wall. 4.5.28 Drumacoo, Galway (M 396 168) Out of the counties in Connacht province, only those churches seen in Galway will be examined on the following pages. Each of the eleven churches to be considered occurs on the same rock type, the Lower Carboniferous limestones of Unit 64 (Table 4.1). All, too, are relatively distant from rocks of different lithologies. A significant geological point should be raised in connection with the more massively bedded varieties of Carboniferous Limestone which tend to be common in Unit 64 which is so prevalent in Galway. This limestone provides much of the pavement or karst type scenery that is widespread in areas like The Burren. The jointing in this rock tends to assist in the creation of the pavement type appearance of the scenery; that is, cracks (joints) tend to incise the pavement surface. This type of readily available limestone provided much of the material for the

Figure 4.88 The south-west anta at Ardagh church displays two courses of ‘Patterned’ stone laid BH and then BVFL, BVFR. Stones at level three are both rebuilt and joggled. (Rule in centimetres).

Figure 4.89 The east wall of St Mel, Ardagh showing the megalithic blocks of face-bedded Lower Carboniferous calcareous sandstone capped more recently with limestone. The calcareous sandstone south-east anta is also shown with the nearest and lowest stone placed BVFR. The plinth is of flaggy slightly calcareous sandstone. 129

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND construction of the earlier churches. In a wall, the jointing or incipient jointing may resemble bedding, particularly when it is viewed from a distance. To distinguish true bedding, structures in the bedding planes, as for example orientated fossils, need to be observed. Determination of bedding orientation tends, therefore, to be more difficult. In the absence of clear bedding orientation, the ‘Patterned’ period stonemason may, presented with these difficulties, have on occasions selected jointing as bedding (although no examples could be confirmed). Reference to the church at Drumacoo has been made by a number of authors and these include Fahey (1901), Hamilton (1903-4), Leask (1958, 73-76), Gwynn and Hadcock (1970, 34), Harbison (1992, 152) and O’Keeffe (2003, 116). A comprehensive account of the church is provided in The Corpus of Romanesque Sculpture (www.crsbi.ac.uk). The church is dedicated to the female saint St Sorney (Sourney, Surney). The early church was subsumed into the larger church that is seen as a ruin today probably about the early 13th century. This was undertaken by extending the church to the south and east so that the original, western trabeated doorway is now considerably off-centre. The earlier church is delineated by the better coursed and more massive masonry in its west and north walls. Both the west doorway and the north-west quoin are constructed in ‘Patterned’style.

Figure 4.90 Drumacoo church (M 396 168) preserves its ‘Patterned’ doorway which is here viewed from the north-west.

Figure 4.91 The west doorway of Drumacoo church as seen from the interior.

130

CHAPTER FOUR

Figure 4.92 The end of the north wall of the original stone Drumacoo church and its north-east quoin is marked by two aligned stones and the wall fabric change. Stones in the flat-headed and splayed doorway are set long and short (or Escomb) style in its jambs (Figures 4.90 and 4.91). There are just three courses in the south jamb set BVFIA, BH, BVFIA. Four courses occur in the north jamb orientated BH (top only visible), BVFIA, BH, BVFIA. In each instance the BH courses consist of two stones, whereas the vertical stones are through stones. The lowest five blocks in the north-west quoin are placed BVFR, BVFR, BVFR, BVFL, BV (direction uncertain) with higher stones in the quoin all replaced. The position of the north-east quoin is evident on the exterior of the north wall (Figure 4.92). The internal dimensions of the original stone church were approximately 7.35m. by 4.54m.

Both nave and chancel preserve some evidence of ‘Patterned’ use of stone. In some respects the church shows characteristics similar to the Anglo-Saxon twocelled church observed so widely in England. In particular, the feature of the chancel being narrower than the nave by the thickness of the wall on the sides of the church is typical. The north quoins of the church appear to have been rebuilt, but both the south-east chancel and south-west nave quoins show vertically orientated bedding in their lowest visible stones: the former; BH, BVFL, BH, BVFL; and the latter, BVFR, BH, BVFL, BH. Only the upper portion of the west doorway is visible (Figures 4.94 and 4.95) but this appears to have been altered, perhaps at the time of the building of a sacristy, to its west, and the construction of the south doorway. Original walls, principally seen in the west, north and south faces are created in large blocks of productid-rich limestone, which is moderately well coursed with the assistance of fairly frequent spalls. The two jambs of the chancel arch are unlike, that on the south side being the older, but modified. The window in the east gable provides the most convincing evidence of ‘Patterned’ workmanship despite being greatly altered internally (Figure 4.96). Externally the north jamb reads BH, BVFIA, BH, BH, BVFIA; and the south jamb BVFIA, BH, BH, BVFIA; the arch is hewn from a single facebedded stone and the relatively massive sill is also facebedded. In the south-east corner of the chancel a small triangular-headed window, although modified externally,

4.5.29 St Cavan, Inisheer, Galway (L 986 025) The Aran Islands off the coast of Clare preserve a remarkable range of early monuments and churches. Despite the impact of early raids and the destruction which they caused, more modern relative isolation has no doubt assisted in the preservation of the ruined churches. There are three main islands, of which Inisheer in the south-east is the smallest and nearest the mainland. St Cavan (Teampall Choemhain) is partially buried in dune sand (Figure 4.93) but preserves both a nave and chancel. Mention or description of this church occurs in a variety of works (as, Westropp, 1895a, 272-273; Mason, 1938; Harbison, 1992, 147) and in the numerous books describing the Aran Islands.

131

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.93 Dune sands cover much of St Cavan church, Inisheer (L 986 025). The church is here seen from the south-east.

Figure 4.94 The west doorway of St Cavan, Inisheer, observed from the east (inside) of the nave. The massive lintel stone of the doorway stands proud and the jambs have been modified. 132

CHAPTER FOUR

Figure 4.95 The west doorway viewed in Figure 4.94 is here seen from the west, the exterior. internally has all stones placed BVFIA in ‘Patterned’ style (Figure 4.97).

Figure 4.96 The exterior of the window in the east gable wall of St Cavan, Inisheer, shows distinct ‘Patterned’ characteristics as may be observed from the stone bedding orientations in the jambs.

Berger (1992, 884) provided a radiocarbon date from this church of 1030-1217. In 1995, from a further sample of mortar in the north nave wall he recorded 1000-1280. Some of this north wall appears to have been rebuilt.

original through stone in each jamb, that in the north jamb BVFIA and, in the south jamb, a BH stone. There is also at least one additional BVFIA stone in each jamb. As detailed here the substantiation for a ‘Patterned’ age for this oratory is limited, cleaning of the east window stones might well supply verification as to the period in which the building was first constructed.

4.5.30 Kilgobnet, Inisheer, Galway (L 975 027) Kilgobnet, Inisheer, is said to be designated to the female saint St Gobnait (St Gobnet). It is a small rectangular oratory and has received only limited description by authors such as Westropp (1895a, 272), Mason (1938) and Harbison (1992, 147).

4.5.31 Teampall Benan, Inishmore, Galway (L 884 071) Temple Benan (Bheanáin) stands near a hill top to the east end of Inishmore, in the Aran Islands (Figure 4.101). The building has been referred to with frequency, particularly on account of its limited size and its unusual orientation. The ‘west’ trabeated doorway faces more to the north, a feature which is probably determined by local topography; whilst the building’s diminutive size, little more than 3 x 2m. internally, provides debate as to whether it represents a church, chapel, oratory or tomb shrine. Reference to the Temple has ranged over authors such as Westropp (1895a, 266), Leask (1955, 49, 53), Harbison (1982; 1992 145) and many others. The site was possibly established by a disciple of St Patrick, St Benen (Benignus), as early as the 5th century, whilst estimates as

Of the standing walls of this church most have been rebuilt as remains evident on the exterior of the east wall where most of the stones bear traces of numbering. In this east gable the small window looks of early character (Figure 4.98), but apart from the face-bedded stone from which the arch has been cut it proved impossible to read the bedding dispositions in its other stones. The eastern quoins are all placed BH. Only the west end of the building shows any indication of ‘Patterned’ features and this only in the lowest stones. The bottom portion of the north-west quoin is constructed of stones placed BVFL, BH, BH, BVFL. The west flat-headed doorway (Figures 4.99 and 4.100) has been partially rebuilt, but it retains an 133

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.98 The east gable window at Kilgobnet, Inisheer, (L 975 027) has been rebuilt as has the east wall (note numbered stones).

Figure 4.97 The triangular-headed window in the southeast of the chancel at St Cavan, Inisheer, internally possesses all the characteristics of a ‘Patterned’ origin. The rule is 325mm. long.

Figure 4.100 A view of the same west doorway at Kilgobnet (Figure 4.99) as seen through the east gable window. The vertically emplaced spacing stones in the fabric of the west wall, observed from a distance only, suggest that this wall may have been rebuilt in ‘Romanesque’ times.

Figure 4.99 The west doorway of the oratory at Kilgobnet, Inisheer, viewed from the exterior. 134

CHAPTER FOUR

Figure 4.101 Temple Benan, Inishmore (L 884 071) viewed from the north-west (the church is orientated with its ‘west’ door towards the north). to the age of the building have ranged from the 8th to the 11th century.

jamb (which is replaced internally), being through stones. The large BH lintel stone is also a through stone.

Almost all aspects of the Temple reflect a ‘Patterned’ origin with only the window in the ‘north’ wall providing a later date with ‘post-Patterned’ stone orientations. (Note: this description will orientate the building as if it were built along an east-west axis, thus this ‘north’ window is more nearly due east). Like the church at St Cavan, Inisheer, the stonework is of a massive productidrich Carboniferous limestone, samples of which are extensively distributed on the local hillside.

Face bedded megalithic stones are used in the walls and the roof clearly had a steep pitch which almost certainly reflected a cover in a material like thatch where the high angle permitted a rapid runoff of water. The walls are relatively thin, 600-680mm. and the solitary window (Figure 4.104), although modified, appears to have originally had an ‘Anglo-Saxon’ keyhole shape. Berger (1995, 165) provided a radiocarbon-dating on the mortar from St Benan and placed the construction date in the range 900 to 1220. The stone bedding orientations probably would confine this date range to the 10th-11th centuries.

A roughly squared plinth, BH, sometimes several courses high, surrounds the building. The four quoins rising from the plinth in megalithic stonework each show ‘Patterned’ style. ‘North-west’ quoin: BVFL, BH, ? (thin stone), BVFL, BH to gable; ‘North-east’ quoin: BVFR, BVFL, BVFR (with graffiti), BH to gable; ‘South-east’ quoin: BVFL, BH, BH, BH, BH to gable; ‘South-west’ quoin: BVFR, BH, BH to gable.

4.5.32 Teampall Bhreacáin, Inishmore, Galway (L 811 121) Temple Bhreacáin (Brecan) has been referred to with frequency in the numerous books which provide information on the Aran Islands. More especially, descriptions by Harbison (1992, 146), who proposed that the west end of the ruined building possibly served as a priest’s house, Westropp (1895a), Waddell (1972-1973) and Manning (2006, 246) should be noted.

The narrow splayed doorway, with four stones in each jamb, displays similar workmanship (Figures 4.102 and 4.103): ‘North’ jamb; BVFIA, BH, BVFIA, BH; and ‘South’ jamb; BVFIA, BH, BVFIA, BH All BVFIA stones, except that in the base of the ‘south’

The off-set, west, blocked and altered, doorway draws immediate attention to the original church which has been extended to both the east and the south (Figure 4.105). 135

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.102 The ‘west’ (actually facing north) doorway at Temple Benan, Inishmore viewed from the exterior. The stonework is noticeably in Escomb style.

Figure 4.103 The ‘west’ doorway of Temple Benan viewed from the inside.

Figure 4.105 The west gable of St Bhreacáin, Inishmore (L 811 121), shows the roof outline of the early church, a blocked and altered west doorway and, the north-west anta in which the stones are set to the ‘Patterned’ style. These extensions result in only the north-west anta being preserved and in its lowest stones a ‘Patterned’ style is revealed. The stones are emplaced: BVFL, BVFR (possibly once a single BVEB stone, now broken, which is cut back on the south side); BVFB; BVFL, BVFR; BH, BVFR with unreadable orientation stone above; probably BH; BH, BVFR: with higher levels missing or replaced. The anta extends 170mm. out from the west wall.

Figure 4.104 The modified window in the ‘north’ wall of Temple Benan viewed from inside. The broken external arching stone is just visible.

The masonry of the original church is massive and distinctly unlike that of the extensions (Figure 4.106), so 136

CHAPTER FOUR that the internal width (3.62m.) and length of the earlier church can be readily determined. A second break in the wall masonry can be observed on the north wall and this marks the later chancel extension (Figure 4.107). On the interior of the early nave extension a triangular-headed window (now blocked) appears to have its jamb stones set BVFIA. This ‘window’ may have been reset, for from its height and appearance in its later life it may have served as an aumbry (Figure 4.108). The wall containing the chancel arch interestingly displays the gable marking of the early extended nave (Figure 4.109). From this evidence it can be concluded that prior to the chancel extension this wall served as the terminal east wall to the nave, there being no visible evidence of a smaller chancel once to the east. The south wall has suffered much alteration and replacement particularly at the south-west corner. A number of medieval ecclesiastical buildings (and an early small ruined church) occur in close proximity to St Bhreacáin, and in several of these the squared and worked stones have been set in post 17th century style with vertically orientated bedding. The same possibly applies to the eastern chancel quoins of St Bhreacáin (although the stones are not worked in the same manner). More probably, the quoin stones represent reused stones from the east end of the earlier church (or its extension) for spalls have been inserted at intervals for levelling purposes. Thus the south-east chancel quoin reads; large BVFR; BH; BVFR; BH; spalls; BVFR; spalls; BVFL; spalls; etc. Figure 4.106 Interior view (that is, from the east) of the west gable of St Bhreacáin. The change in the masonry fabrics between the early church and its southern extension is very evident.

In summary, this church which was originally unicameral, possessed antae and was built in the ‘Patterned’ period. It was extended eastwards (possibly prior to the ‘Romanesque’ period), and then on the south side to widen the church. Evidence that the chancel extension followed later can be seen by the wall fabric join low in the south wall (Figure 4.110).

Figure 4.107 Two masonry breaks occur in the north wall of St Bhreacáin viewed here from the north-west. The more pronounced occurring between the extended nave and the chancel. 137

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.108 In the extended north wall of St Bhreacháin viewed from the south, Was this ‘aumbry’ once a ‘window’?

Figure 4.109 The visible gable markings of the extended nave above and to the south of the chancel arch at St Bhreacháin.

Figure 4.110 The growth of St Bhreacháin, Inishmore. The present building is sketched on the left with dotted outlines of the positions of earlier walls. The outlines are intentionally not to scale and arrows indicate positions of clear wall fabric change. Successive stages in the building growth are also shown. a = west doorway in use, a* = west doorway blocked; b = south doorway; c = early window, c* = early window blocked, to become aumbry; d = anta, when present; e = chancel arch. 138

CHAPTER FOUR

Figure 4.111 View towards the south-east interior corner of Temple Chiaráin (L 873 104). The east wall window and its encompassing elaborate string course are likely to be of an early 13th century ‘Romanesque’ date (see text). 4.5.33 Teampall Chiaráin, Inishmore, Galway (L 873 104)

4.5.34 Teampall MacDuagh, Inishmore, Galway (L 823 104)

Much the same range of authors has referred to Temple Chiaráin as for previous Aran Island sites. One book giving a general and modern guide to the Aran Island ecclesiastical sites is that by Jones (2005).

It has been suggested that Temple MacDuagh was founded by St Colman in the 7th century. Certainly, the western end of the current ruin, with its massive masonry looks old (Figure 4.112). The original church clearly occupied the western portion of the ruin and here it still retains the antae, both of which exhibit their ‘Patterned’ origin. These antae extend about 330mm. from the west wall face. They read: North-west anta: BVFL, BVFR (possibly one stone now split); BH; BVFL and BH upon BH; BVFR, BVFR; BH; BVFR upon BH and BVFR; with higher stones rebuilt. South-west anta: BVFL, BVFR; BH; unable to determine, BH; BVFB; Unable to determine, south stone BVFR; BH, BH. The lowest three stones that now form each of the eastern quoins of the nave (the remnant antae) also show ‘Patterned’ orientations.

This much altered ruin was dedicated to St Kieran (Harbison, 1992, 146). Single-celled, it originally had a west doorway which is now extensively modified. That west entrance was no doubt replaced at some period by a more modern doorway in the north wall. Little of any of the original church remains (Figure 4.111), but the window in the south wall (in all probability once matched by a similar window in the north wall) is probably early ‘Romanesque’ in origin, with its jamb stones set BH and the arch stone typically BVEIA. The single east window is virtually identical in style to those that comprise a pair in the east chancel wall of O’Heyne’s church at Kilmacduagh (M 404 003), which is regarded as having been created in the late 1220s to mid 1230s and being workmanship of the ‘Romanesque’, ‘School of the West’ (O’Keeffe, 2003, 115).

The west doorway, which is both splayed and flatheaded, also possesses ‘Patterned’ characteristics. Each of the jambs consists of large stones all laid BH, but for one stone in each orientated BVFIA. The megalithic lintel stone is of granite, possibly gathered as a glacial erratic boulder on the island. The trace of a cross may be seen carved above this stone.

Much as in the instance of Temple Cronan church, in the Burren, each of the four quoins rises to a corbel. These corbel stones are bedded vertically. Each of the quoins has been modified and repaired but the south-west quoin still retains three stones, of the twelve, placed with vertically orientated bedding as well as a BVFR corbel stone. Such proof of a ‘Patterned’ origin must remain, therefore, somewhat tentative.

The chancel built on to the east end of the original building shows no evidence of early origins. Similarities in masonry styles in the lower parts of the respective south walls of the current nave and chancel (Figure 4.113) no doubt reflect re-use of earlier nave east gable wall material. Harbison (1982, 621) recorded the 139

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.112 Temple MacDuagh, Inishmore (L 823 104) is viewed here from the north-west to show the western antae and doorway.

Figure 4.113 The south walls of the chancel (right) and the older nave of Temple MacDuagh display similar style masonry. This is probably because the stones in the lower part of the later chancel include re-used massive stones from the east end of the earlier nave. 140

CHAPTER FOUR

Figure 4.114 Little still stands of Temple na Neeve, Inishmore (L 824 103). The orientation of the bedding in the lowest four stones in the south-west quoin, nearest the camera, suggests that the church has a ‘Patterned’ origin. presence of a carving of a horse on the north church wall, which he compared with a similar carving on the early 13th century church at Inishmaine, County Mayo (M 138 617). Berger (1995) provided a radiocarbon date of 10701300 from mortar from the east wall beside the south-east quoin. This date would relate well to that of the later chancel extension.

‘Romanesque’ times, for some of the masonry shows narrow fillets of vertical bedded stone set into spaces in courses, typical of that period. At the north-west corner the stones are more massive and the quoin stands higher (Figure 4.115). Two small stones have been inserted at the base of the quoin as replacements but above this the megalithic stones read, in ‘Patterned’ style, as: BVFL; BVFR; BVFL; BVFR; BH; BH; followed by replacement stones.

4.5.35 Teampall na Neeve (Naomh), Inishmore, Galway (L 824 103) This small ruined church is also known as the ‘Church of the Saints’. Little of the church now stands. That the original building possessed cyclopean masonry and a western door is evident. The eastern end of the ruins shows signs of having been rebuilt, the south-east quoin is missing, but the stones in the remaining north-east quoin are all laid BH. The south-west quoin rises in four stones to about 2m. and these are set; BH; BH; BVFL; BVFR; replaced stone (Figure 4.114); from which a ‘Patterned’ origin might presumably be inferred.

4.5.37 Kilmacduagh Cathedral, Galway (M 405 000) This ruined cathedral on mainland Galway forms part of a monastic site (Figure 4.116) which is believed to have been founded by St Colman, son of Duagh, in the early part of the 7th century (Harbison, 1992, 158). Further references related to the building include Fahey (1904), Leask (1960), Harbison (1982) and Manning (2006). The site is established on Lower Carboniferous limestone (Unit 64, Table 4.1) and this rock type provides the building stone for the church.

4.5.36 Temple Soorney, Inishmore, Galway (L 866 105) The quoins at the western end of the nave preserve their ‘Patterned’ character. Rising above the plinth the northwest quoin displays stones with the following bedding orientations: undetermined, BVFR, BH, BVFR, BH, BH, followed by three BH more recently inserted stones and a BH corbel. The south-west quoin exhibits: BVFL (partly repaired), undetermined, BH, BH, BVFL, BH, followed by four more recently inserted BH stones and a corbel. Examining the north and south walls (Figure 4.117) it is

Temple Soorney is also described as Teampull Assurnidhe (Assurnuidhe). It is now in a very ruinous state with the original west doorway in a wall just a few stones high. The lateral walls show elements of outward curvature and the east wall rises only to the base of a onetime east window. The building has probably been largely rebuilt for three of the quoins have all remaining stones set BH. That rebuilding may have occurred in 141

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.115 The ruins of Temple Soorney, Inishmore (L 866 105) viewed from the south-west. The nearest corner exhibits stones set in ‘Patterned’ style.

Figure 4.116 The monastic site of Kilmacduagh (M 405 000) viewed from the nearby road. Beside the round tower is the cathedral with the ruined church of St John the Baptist nearer to the viewpoint.

142

CHAPTER FOUR

Figure 4.117 The north wall of the nave of Kilmacduagh Cathedral clearly shows the fabric change and wall building that has been undertaken above the height of the north-west corbel. clear that both the height of the gable and the side walls have been increased above the level of the corbels (possibly in the 15th century). There is a gap between the inner and outer west gable walls at this height (Figure 4.118) and this gap probably was initially filled with loose unmortared rubble which has settled and fallen away (see Fernie, 2000, 295 and section 2.13.3). Workmanship of this nature is known from various Norman and later churches in England.

much extended eastwards subsequently added.

and

transepts

were

4.5.38 Kiltiernan (Cill Tiarain), Galway (M 437 156) Reference to the ruined Kiltiernan church has been made by a number of authors, including Westropp (1919, 178), Duignan (1951), Leask (1955, 52, 67), Gywnn and Hadcock (1970, 395), Harbison (1992, 160), Waddell and Clyne (1995) and O’Keeffe (2003, 71). It has been suggested that the church site is one upon which St Tiernan established a monastic site about 550. The site is on Lower Carboniferous limestone (Unit 64, Table 4.1) and the church is built of this material.

The partially blocked west doorway to the nave also provides evidence to the early origin of the west end to the cathedral. Flat-headed, splayed, and built of megalithic stones, the doorway displays, as far as can be determined, ‘Patterned’ characteristics (Figures 4.119 and 120). The jamb stones carry both lichen growth and a limewash coating so that bedding orientation is difficult to resolve. The partial blocking also makes the relationship between the external and internal aspects of the stones complicated and which are through stones in the lower portions remains uncertain. The top two stones in each jamb are probably through stones (the lower of the two in the north jamb being proposed somewhat tentatively) and the top stone in each is placed BVFIA and that below BH. An early square plinth exists beneath the west gable wall. Above the top of the doorway much of the gable wall appears to have been rebuilt at least on the exterior. It also shows that an earlier roof pitch was steeper (Manning, 2006, Fig. 4). The cathedral has been

The early, much repaired, church has been extended to the east to create a chancel, so that only the west end of the nave preserves projecting original antae. A single, and sometimes double, early square plinth surrounds the earlier nave and on the south side of the church this returns beneath the later chancel wall. The west doorway takes the customary early form in being flat-headed, splayed and built of megalithic blocks (Figures 4.121 and 4.122). Both the heavy lintel and the original jamb stones are through stones, except where the south jamb has been rebuilt on the inside. Externally, three large stones each set in ‘Patterned’ BVFIA style create the north jamb, with the principal stones in the south jamb being BVFIA, BVFIA, uncertain through stone. Above the western doorway smaller wall fabric blocks indicate rebuilding. 143

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.118 A view from the north of the west gable of the nave of Kilmacduagh Cathedral. The heights of the gable and the side walls have been raised, probably in ‘Romanesque’ times. Note the gap between the inner and outer gable wall surfaces.

Figure 4.119 The west doorway of Kilmacduagh Cathedral nave as observed from the west.

Figure 4.121 The west face of the ruined church at Kiltiernan, Galway (M 437 156). The change in masonry fabric above the level of the top of the lintelled doorway should be noted.

Figure 4.120 The nave west doorway of Kilmacduagh Cathedral viewed from the interior (east). Compare with Figure 4.119.

144

CHAPTER FOUR

Figure 4.122 Inside view (from the east) of the ‘Patterned’ west doorway at Kiltiernan church, with cow for scale. Each of the western antae is placed on a double squared plinth. Little of the south-west anta is preserved (the south stone on the lowest course is set BVFR, Figure 4.123), but the lowest stones in the repaired north-west anta are placed as follows: first course, BVFL, ?, BVFR; second course single stone uncertain; third course, BH and ? (repaired); fourth course, repaired and BVFL; with the higher remaining portion of the anta restored. Traces of both original eastern antae (with a number of vertically orientated stones) prove to be visible where the later chancel has been built to adjoin them and the original church (nave) east wall removed (Figure 4.124). Here, it can be observed that both the original nave and the later chancel were of thicker wall structure than the antae.

surrounding area includes Dalradian schists and phyllites and Palaeocene basalts. All of these rock types occur in the walls, but in addition cobbles and boulders of granite and granodiorite are present, these probably initially transported by streams from Mullaghmore to the south. Amongst others, the following make important contributions to the study of Banagher church; Champneys (1910, 228), Leask (1955, 83), Waterman and Hamlin (1976), Hamlin (1984, 122), Harbison (1992, 89). The last of these authors indicated that the church was supposed to be founded by St O’ Heney whose tomb, close by, is supposed to date from 1210-1225 (but looks more modern). Harbison, therefore, wrote, ‘the church is scarcely any older than the 12th century’.

In the south wall there exists an interesting example of what is probably a ‘Romanesque’ triangular-headed window (Figure 4.125). In England, windows of this shape, although not common, are typically regarded as being of relatively late Anglo-Saxon origin. There, as far as can be determined, they possess jamb stones in which the bedding is placed vertically. At Kiltiernan, the jamb stones, although often again through stones, are set BH. Other Irish examples have been observed with these stone settings, elsewhere as at Trinity church, Glendalough (T 128 968) and Kilmacduagh, St John the Baptist (M 405 001). The walling in the chancel is of different style to that in the earlier nave.

Banagher, like other churches discussed above, has been extended towards the east to create a chancel. The west doorway, to the nave and the original church, is splayed and possesses a bold lintel (Figure 4.126). The whole is, however, encompassed with an elaborate but worn architrave and internally the lintel is replaced by a round arch. The jamb stones are not in ‘Patterned’ style and with one exception its regular stonework is set BH. Waterman and Hamlin (1976) related this doorway on its style to that, for instance, at Aghowle in Wicklow (S 932 694) which Leask (1955, 85) referred to a date of 1100. Although all stones in the east chancel quoins are emplaced in orthodox BH fashion, the stones in the lower parts of the four quoins of the nave, the original church, display ‘Patterned’ characteristics, as follows: North-west quoin; plinth, BVFL, BH, BH, BH, BH, BVFR, probably BVFL, remainder BH. South-west quoin; plinth, BVFL, BH, BH, BH, BH, BVFR, BH, BH, BVFR.

4.5.39 Banagher, Derry (C 676 066) Just one example of a church from Ulster that appears to preserve remnants of its ‘Patterned’ origin is described here. Typically, Banagher church ruin reflects the complexity of the geology in the province, for its walls consist of a variety of rock types. Underlain by Lower Carboniferous sandstones (Unit 59, Table 4.1) the 145

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.123 The south-west anta at Kiltiernan ruined church and the double plinth upon which it rests.

Figure 4.124 The south-east anta at Kiltiernan viewed from the north (inside) towards the point where the gable wall once joined the south wall (right). Note that the nave south wall is thicker than the anta wall. (Rule 325mm. long).

Figure 4.125 The triangular-headed south nave wall window at Kiltiernan church. This window is believed to be of early ‘Romanesque’ age having been inserted at the time of the chancel extension and the loss of the east window in the nave.

Figure 4.126 The west doorway to Banagher church, Derry (C 676 066) seen from the exterior. Note that an architrave surrounds the whole doorway and that the stonework is set in BH style. 146

CHAPTER FOUR

Figure 4.127 The nave plinth, of square section, at the south-east corner of the nave of Banagher church turns beneath the later chancel wall, which possesses a ‘Romanesque’ style plinth. (Rule 325mm. long). South-east quoin; plinth, uncertain, uncertain, BH, BH, BH, BVFR, remainder replaced BH. North-east quoin; plinth, BVFR, BVFL, remainder replaced BH. All the stones in the ‘Patterned’ portions of the quoins were of Carboniferous sandstone.

elaboration is now provided. The sandstone plinth occurs beneath all walls, but the plinth on the east wall of the nave noticeably turns underneath the later chancel walls (Figure 4.127). The plinth beneath the west and south nave walls is square in section whereas below the chancel it is of ‘Romanesque’ style. Beneath the north wall of the nave the plinth was probably modified in ‘Romanesque’ times.

Waterman and Hamlin (1976, 29) noted the variation in the styles of wall fabric in the church ruin, and some

Figure 4.128 Three major variants in wall fabric style are exhibited in this view of the north face of Banagher church viewed from the north-west. They occur principally; low in the nave wall, higher in the nave wall and in the more distant newer chancel. See text for further details. 147

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.129 A view of the chancel fabric and south chancel ‘Romanesque’ window at Banagher church as seen from the south-west. Externally, three principal wall fabric varieties may be distinguished (Figure 4.128): In the lower portions of the nave wall: mainly sandstone and Dalradian slate and phyllite (sometimes as boulders), but with included cobbles and boulders of conglomerates, and basalt, granodiorite and other igneous rocks (especially within the north wall); these are moderately coursed with few spalls. In the higher courses of the nave wall, representing rebuilding: sandstone in more parallel-sided pieces is poorly coursed with spalls uncommon. In the later chancel; Squared and chiselled sandstone in courses of rectangular blocks occur (Figure 4.129). Inside the chancel, bands of Dalradian phyllite and slate are added untidily to the sandstone courses for levelling purposes suggesting that the interior was intended to be plastered. However, at the western end of the south wall of the nave, for the lowest two metres the fabric is only of sandstone, occasionally as boulders. Whether this represents very early walling or more probably a later repair (possibly infilling an earlier south doorway), is uncertain The distribution is somewhat more complex than it is possible to describe briefly.

examined there were other sites in which limited elements of ‘Patterned’ workmanship could possibly have been identified but, typically they involved instances where the stonework was largely unreadable or only a very small number of stones were emplaced with vertical orientation. Even within the 38 listed above, in some cases, doubts in interpretation remain, although given more detailed examination and possibly certain stone cleaning, clear conclusions as to the style of workmanship might be made. Removing three sites where elucidation might be questioned; Nougaval (possibly an old mausoleum), Kilgarvan (more study required) and Temple Chiaráin (limited stonework), 35 churches remain for consideration. Others might be more selective in their choice. The 35 churches fall into seven broad categories: a). Simple, small unicameral churches or oratories (11 sites). b). Unicameral churches, again of small size, built with antae (4 sites). c). Type a) extended to the east to provide a chancel (8 sites). d). Type b) extended to the east to provide a chancel (9 sites). e). Type a) extended to the west to provide a nave (1 site). f). Extended to the west (Agha) (1 site). g). Possibly built as a two chambered (bicameral) church (1 site).

4.6 Brief analysis of Irish churches described which possess ‘Patterned’ characteristics In section 4.5, thirty-eight church sites have been briefly described. In the much larger number of churches 148

CHAPTER FOUR Both churches involved in extension to the west; that is Ardmore (Type e.) and Agha (Type f.), were probably constructed in this way because of topographical restrictions, the land falling away steeply in the first instance, and rising in the second, towards the east.

BVFB – BH courses (occasionally including BVEB stones in the patternwork). With regard to window arches (e), insufficient windows were examined in this analysis to assess any similarities in the stone bedding orientations in the heads of other forms of window (such as rubble or triangular-headed varieties) of the ‘Patterned’ period. Similarly, with only the south jamb of the early Gothic style chancel arch at St Cavan, Inisheer providing any evidence of ‘Patterned’ stonework, it is difficult to confirm that all pre‘Romanesque’ chancel jambs in Ireland followed the practice of England and Scotland in being of the BVFIA – BH style.

In only one instance of the 35 churches is the chancel thought to display certain datable ‘Patterned’ characteristics (category g), namely St Cavan, Inisheer. In the instance of Friar’s Island (Killaloe), the chancel probably post-dates the nave, and the church is placed in category c). Several churches were extended in more than one direction: Ardmore (additionally to the east), Glendalough Cathedral (if the arguments of Manning (2002) are followed) and Drumacoo and Temple Bhreacáin (both to the south).

The identification of face bedded stones within the face of masonry walls, as emphasised in churches both in England and Scotland where they have been observed, can be exceptionally difficult. Only one surface of the stone is visible, whereas usually in a quoin, jamb stone or anta two surfaces can be scrutinised for bedding orientation. In the often massively bedded and finegrained rock lithologies from the Carboniferous Limestone of south-west Ireland a flat surface may be difficult to identify. Particular care has to be taken in distinguishing between bedding, jointing and cleavage, in rocks which may well be covered in a millennium of supplemented grime and lichen. Other church wall masonry faces, therefore, may hide similar clear patterns to those observed at Glendalough Cathedral. A singleband of face bedded stones can, for instance, be traced round much of the external walls of St Mochta’s House, Louth (H 956 014) at 1.5m. above ground level, suggesting the fabric of the lower walls may be of ‘Patterned’ origin.

Churches in categories c), d) and e) frequently provide evidence that their extension occurred during the ‘Romanesque’ period. The newer area in each of these churches lacked any evidence of ‘Patterned’ workmanship and only exhibited features of ‘Romanesque’ or later styles of architecture. In certain cases the ‘Romanesque’ work in the newer extension has been tentatively dated by others; examples being: Ratass, ‘shortly after 1111’ (O’Keeffe, 1998, 121); Ardpatrick, ‘probably around 1200’ (Harbison, 1992, 216); Glendalough Cathedral, ‘c. 1200’ (Manning, 2002) and Drumacoo, ‘early 1200s’ (O’Keeffe, 2003, 116). ‘Romanesque’ work may also be inserted into the earliest church walls, for example to provide additional light, as in the south wall of Kiltiernan church. The critical message from these observations, however, is to note that the ‘Patterned’ areas of the buildings always pre-date those of ‘Romanesque’ age.

4.7.2 Architectural features of the ‘Patterned’ period

4.7 The ‘Patterned’ characteristics of early Irish ecclesiastical buildings

Just as the rock types of Ireland more closely resemble those of Scotland and northern England, so the architectural characteristics of the ‘Patterned’ style Irish churches more nearly approach those of the same regions. The features of the period identified may be listed as:

A range of stylistic features which can be used to distinguish ‘Patterned’ buildings from those of later buildings in Ireland can now be itemized, this being based on the evidence provided from section 4.5. These may be classified into two broad groups: those related to bedding in stonework that is vertically orientated, and those exhibiting a variety of architectural characteristics.

a). Wall masonry may show certain distinctive qualities. i. If a source of suitable cobbles or boulders (river, beach, glacial, fluvioglacial) is available, material from this will probably have been used in preference to hewn stone. On such occasions the area of mortar between stones is greater and the stones may have many origins. Low in the walls of both Banagher and Kells this type of masonry is exemplified. ii. Other than the wall being of a composition as above, it will generally have been extracted from only one local source. iii. Large, megalithic stones were popular and where used in wall masonry they tended to be placed in the face bedded situation. Reference to the use of such stones will be made in sections 5.2 and 6.3.2. Such stones could of course be re-used in rebuilding. b). As indicated by Taylor and Taylor (1965, 13) in England, plain square section plinths, especially

4.7.1 Vertically orientated bedding in stonework a). In the quoin stones of wall corners: mixed BVFR, BVFL, BH stonework. b). In wall antae: BVFR, BVFL, BVFB, BVEB, BH stonework. c). In door jamb stones: BVFIA and BH stonework, often set to Escomb style. d). In window jamb stones: BVFIA and BH stonework (as at St Cavan, Inisheer). e). Round headed, single-splayed windows with single arch stones typically set BVEIA. f). On occasions where suitable well bedded or cleaved rocks were available (as at Glendalough Cathedral), the introduction of patterned wall ornamentation with 149

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND where constructed of large flat stones, tend to be distinctive of the period. c). The use of through stones in openings such as for doors and windows was typical of the period. d) Tapered and flat-headed (trabeated) doorways, as in Scotland, were part of the ‘Patterned architectural style. Such doorways were not particularly common in Scotland, however, in Ireland they were extremely popular and possibly widespread e). Many of these doorways were provided with an architrave created by cutting back the framing stones, as observed particularly in England (Potter, 2006c). The width of the cut back was again determined (and in the same manner) by the smallest stone. In England, architraves of this type were probably only created post about 950 and it would seem possible the same might be true for Ireland. f). Of those ecclesiastical buildings listed in section 4.5, almost all possess or had probably possessed a western doorway, these entrances, in many instances being used until the time the church fell into disrepair. In England, the western entry was used, particularly in the older ‘Patterned’ churches, but was frequently replaced with north and south Post-Conquest doorways. The continental situation, where the west entry predominates (Taylor, 1978, 831) is more like that of Ireland. g). There is evidence, that as in Anglo-Saxon workmanship in England, ‘Patterned’ walls were constructed to be thinner than ‘Romanesque’ walls, as is apparent at Ardpatrick. Figure 4.130 The chapel built by Cormac Mac Cárrthaig (MacCarthy) at Cashel (S 075 409) viewed from the south-east. Clearly of ‘Romanesque’ style, the chapel has many similarities with the church at Bradford on Avon (see Figure 2.68).

4.7.3 Other non-definitive factors of the Irish ‘Patterned’ period It had been hoped to comprehensively tabulate the factors detailed in sections 4.7.1 and 4.7.2 and to compare them with the Scottish and various English ‘Patterned’ practices. The exercise proved too complex because of the many variables and referral must instead be made to sections 2.4 and 2.7 and Table 3.1. The evidence listed, however, certainly emphasizes the similarities in ‘Patterned’ stonework traditions found in the three countries and a simplified comparison is provided in Table 4.3. Not all aspects of similarity between the workmanship of the period have necessarily been covered, it being more appropriate to discuss some at the point where they first arise in the text.

4.7.4 The Irish ‘Patterned’ period: its duration Attempts to determine the date of commencement of the ‘Patterned’ style of ecclesiastical building in Ireland is fraught with the same difficulties as in England and Scotland: the lack of explicit documentary evidence as to the date of construction of the buildings (see section 2.9.2). Vertically bedded stonework can be observed in what are thought to be the earliest stone buildings in each country. In England, buildings like Bradwell-on-Sea or the chancel at St Martin, Canterbury (TR 158 577), thanks to the writings of Bede, take this date back to the early 7th century. In Ireland, the date for the earliest stone church (or damliac) tends to be regarded as somewhat later (see section 4.1).

Certain differences also occur. Most of these, like the absence of pilaster-strips in Ireland, can be directly related to variations in the available building materials, that is, the rock type. The relative ease with which the younger stone types in England could be extracted and worked, might possibly account for height of the church walls in England as compared with most in Scotland and virtually all in Ireland. An alternative possibility might be that the softer climate in England with less strong winds permitted the masons there to build their walls to a greater height! The outstanding difference, the presence of antae in Ireland, will be discussed in Chapter 6.

The resolution of the date of the replacement of ‘Patterned’ by ‘Romanesque’ buildings in Ireland presents similar difficulties. For many, the chapel erected by Cormac at Cashel (S 075 409) represents the earliest truly ‘Romanesque’ building in Ireland (de Paor, 1967). The chapel (Figure 4.130), constructed over the period 1127-1134 (Garton, 1981; Harbison, 1992) is certainly 150

CHAPTER FOUR Table 4.3 A simplified assessment of some of the stonework features which are believed to distinguish churches of the ‘Patterned’ (in England, Anglo-Saxon) style. This assessment applies to early churches in Britain and Ireland and in all instances is influenced by the availability of suitable rock lithologies. The information given here is based upon, and supplements, the details provided in Potter (2008c; 2008d) Architectural or Stonework Feature A) Long recognized features Long and short quoins (geometrical orientation) Megalithic quoins Double-splayed windows ‘Escomb-style’ doorways (geometrical orientation) Pilasters, pilaster-strips Other features (e.g. string courses, plinths, etc.) B) Newly observed features Vertically bedded stonework set to specific styles in quoins and arch jambs Vertically bedded stonework set in pilasters Cut back stonework to produce apparent equality of width Cut back stonework in doorway architraves Polychrome banding Face-bedded banding Megalithic face-bedded stones

Exemplified particularly in parts of

Typical Rock Lithology

England

Well bedded, moderately massive, easily worked limestones and sandstones Massive, difficult to work, harder rock types, in instances the re-use of Roman stones Easily workable rubble (rare exceptions) Similar to long and short quoins

Northern England, but depends on rock types South and East England England England (not extreme north) (See Taylor & Taylor, 1965)

Similar to long and short quoins

Wherever rock type suitable England

Well bedded or lineated, workable, moderately massive stone Well bedded, easily worked stone, particularly limestone Similar to pilasters

England Ireland, Scotland East Anglia and South-east England Scotland, Ireland, northern England Scotland, Ireland

Varies according to feature

Stone of massive and sound quality Rubble church walls, particularly where Superficial deposits of different colour Very well bedded or lineated rock types as fine grained sandstones to work as ashlar As face-bedded bands, but massive, and often more slaty rock types

Figure 4.131 Gallarus oratory (Q 393 048), viewed here from the west, displays no ‘Patterned’ workmanship. 151

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 4.132 Temple Cashel oratory (V 369 687) also viewed from the west was many similarities with Gallarus. devoid of ‘Patterned’ stonework. As noted in section 4.6, O’Keeffe (1998; 2003) has dated certain Irish ‘Romanesque’ structures in buildings to around 1110 and by that date possibly the new fashion in architecture had swept in to Ireland (see sections 2.8 and 4.1). 4.8 Ecclesiastical Buildings of the Gallarus Type Clocháns or beehive huts, built with corbelled stone roofs, have been constructed in Ireland since the Bronze Age, and quite naturally, oratories built to a similar roofing style have until moderately recent times been unquestionably regarded as of particularly early pre‘Romanesque’ style. Leask (1955, 27 et seq.) regarded an oratory such as Gallarus, Kerry (Q 393 048) as evidence of an early structural style in the evolution of the Irish church (Figure 4.131). Harbison (1970), however, offered strong arguments for suggesting that Gallarus, and possibly other oratories of the same construction, were not built until as late as the 12th century. Subsequently, the debate as to the age of these corbelled oratories has continued (see Hare and Hamlin, 1986). Following the excavation of the oratory at Illaunloughan (V 365 731) (White Marshall and Walsh, 1998; 2005), Rourke and White Marshall (2005) provided an excellent analysis of corbelled oratories and proposed an evolutionary progression in their design. Illaunloughan they concluded was a structure built early in this evolution with a radiocarbon date broadly set on the eighth century. Vertically bedded stones are noticeably of frequent occurrence on the site, particularly within the structure of the integrated leacht.

Figure 4.133 Detail of the east window at Temple Cashel. Note that all stones are laid BH and that the window possesses both a protective lintel and a sill. To the right of the pen (135mm. in length) a hole has been drilled through the lintel stone for fastening the hinge of the window. 152

CHAPTER FOUR It should be noted that of the better preserved corbelled oratories examined by the present author, in buildings where each roof stone is individually shaped, such as Gallarus and Temple Cashel (V 369 687) there is no indication of vertically bedded stone emplacement of a style typical of ‘Patterned’ craftsmanship. Gallarus includes one stone in the north jamb of the door set in 12th century or later, ‘Romanesque’ BVEIA style, but all other visible rocks in the oratories of this type which were examined appeared to be otherwise set with their bedding horizontal. Based on their analysis of the range of corbelled oratories, Rourke and White Marshall (2005, 119-120) concluded that Gallarus was the most sophisticated structurally and represented the final phase of development in these buildings. They considered Temple Cashel to be akin to Gallarus. In Temple Cashel (Figure 4.132), for instance, a protective lintel extends over both the doorway and the east window (Figure 4.133). Above and below the window there are drilled holes to hold a hinge and fastenings. The Rourke and White Marshall analysis, together with the evidence from the stonework would, therefore, appear to fully support Harbison’s 12th century date. Other corbelled oratory sites examined such as Reask (Q 367 045), excavated by Fanning (1981), and similar early Christian sites such as Killabuonia (V 403 698) and Lough Gur (R 646 415) noticeably include vertically emplaced stones. It should be noted that the record from England of Anglo-Saxon vertical stone emplacement extends back to include the earliest Anglo-Saxon stone churches (Potter, 2005b).

The ‘Patterned’ use of vertically bedded stones assists in churches of this age being differentiated from those ecclesiastical buildings of later periods. As a result, those corbelled buildings of the later Gallarus type are thought to be of no earlier date than the 12th century. With this pre-‘Romanesque’ fashion of stone emplacement providing a new marker to support the identification of building dates, the periods of use of other architectural features may be distinguished more readily. For instance, architraves to doorways are a late Anglo-Saxon feature in England (Potter, 2006c). In Ireland, this ornamentation appears to continue into the early ‘Romanesque’ period, particularly in certain round tower doorways. The reason for the absence of Anglo-Saxon pilasters in Ireland has been attributed to the different nature of the rocks available for use. The possible rationale for antae being restricted to Ireland will be described in Chapter 6. The suggestion that stonemasons working in the ‘Patterned’ style, as observed at Glendalough Cathedral, on occasions placed stones in masonry courses with the bedding (or in the case of Glendalough, cleavage) vertical and in patterns, has important implications for those involved in church restoration. When any stone is removed from a wall, the stone orientation should be recorded. At the time of replacement the stone should be reset precisely to follow the original orientation.

The question that immediately arises from the conclusion of accepting a 12th century date for Gallarus style oratories, is why in Ireland should this type of structure be erected at a time when ornate and truly ‘Romanesque’ churches were being built (Rourke and White Marshall, 2005)? The answer must surely relate at least in part to the geology. The distribution of corbelled roof buildings is such that they are virtually confined to the barren topography of south-west Ireland, a wind-swept region of hard Palaeozoic sandstones, where rocks for building purposes are plentiful but trees are scarce. The foremost demand for timber would have been be for fuel, with large timber suitable for roofing in short supply. Well constructed stone roofs were an obvious alternative. 4.9 Conclusions and implications The ‘Patterned’ fashion of placing stones with vertically bedded orientation into certain structural features of ecclesiastical buildings has been demonstrated as occurring in Ireland, as elsewhere. No doubt more buildings will in time reveal the same characteristic stone emplacement details. It is not proposed that this vertical stone setting was the result of an influx of Anglo-Saxon masons who executed this work. Instead, it resulted from Irish craftsmen following, or possibly developing, aspects of a widespread fashion in the same manner as those in England and Scotland. The region from which the styles originally developed has yet to be determined, but the movements of early missionary monks and pilgrims must have greatly assisted their spread. 153

CHAPTER FIVE A CRITIQUE OF THE GEOLOGICAL ASPECTS OF THE SUPPOSED MONASTIC INFLUENCES IMPOSED ON EARLY IRISH ECCLESIASTICAL MASONRY STYLES Carragáin (2005a). He believed that he could identify five distinct patterns. Ní Ghabhláin (1995), working on a single site, in contrast, identified only three variants which she related to different periods of building. In contrast, Leask (1955, 52) believed that masonry styles could not independently define construction dates of Irish churches. He, like many other authors, such as Petrie (1845) and Manning (1995a, 16), however, assigned very large, cyclopean stonework to a broad, but specific ‘preRomanesque’ period.

5.1 Introduction As indicated in section 4.1, several authors have attempted to analyse the various forms of masonry which occur in early Irish ecclesiastical buildings. Recently, detailed analyses of the differences in wall masonry styles of churches have been attempted (Ní Ghabhláin 1995; Ó Carragáin 2005a), that of the latter author being a particularly detailed and painstaking analysis. Ó Carragáin concluded that different observed wall masonry styles were ‘essentially habitual’ (2005a, 143) and reflected the work of regional or ‘local groups of masons that ….expressed themselves most clearly in the very fabric of the churches that they built’ (2005a, 144). This theme of regional variation, founded on a classification of four Irish early church types proposed by Harbison (1982), was further developed by Ó Carragáin where he stated, ‘A number of factors, both archaeological and historical, combine to suggest that the local styles developed in the period from the latter half of the eleventh century to the early twelfth century’ (2005b, 32).

The geological evidence, in contrast, would suggest that different masonry styles are largely determined by the significant differences in rock lithology (Figures 5.1 to 5.4). Confusion tends to occur as soon as walls of different rock types are compared. Rock types should be examined individually in order to establish the manner in which methods of working particular stone lithologies changed with successive periods (as Potter, 1998). Many rock types, for instance, could never have provided early stonemasons with material sufficiently massive and appropriately bedded to create megalithic stonework. Furthermore, there is certainly no evidence to suggest that these masons travelled long distances to obtain the cyclopean stones.

Contrasting early styles of wall fabric masonry can certainly be distinguished and are well displayed by Ó

Figure 5.1 The inside of the south wall of the older ruined church at Maghera, Co. Down (J 372 342) exhibits an interesting collection of local rock types which include blocks of columnar basalt, boulders of granite and spalls of altered greywacke.

Figure 5.2 The Burren Carboniferous limestones (Unit 64, Table 4.1), seen here in the walls of Oughtmama East church (M 305 079) as viewed from the south-west, provide distinctly different wall fabric to that seen in Figure 5.1.

154

CHAPTER FIVE

Figure 5.3 Walls of partly hammered and shaped blocks of Upper Old Red Sandstone (Unit 54, Table 4.1), seen here in the ruined oratory at Killelton, Kerry (Q 720 100), afford a further wall fabric to those seen in Figures 5.1 and 5.2.

Figure 5.4 Viewed from the west, St Kieran, Glendalough (T 123 967) is built of local phyllite which provides a different wall fabric again to those seen in Figures 5.1 to 5.3. Note that the lower stones in the doorway jambs are orientated vertically.

Ó Carragáin (2005a) chose to carefully photograph, to the same scale, portions of the masonry exposed in the lower walls of a large number of Irish churches that he regarded as being of an early date. No record of which face of the building and whether the masonry was inside or outside the building was detailed. As such this could mean that areas of wall that had been rebuilt (and perhaps relatively recently) could have provided the image (Figures 5.5 and 5.6). Instances where the masonry differs internally and externally also occur (See Figures 4.23 and 4.24, and 5.7 to 5.9). The Ó Carragáin photographs were apparently used for the purposes of comparative analysis using nine different visual characteristics of the masonry. In all, 195 churches were tabulated although possibly only 149 were found to be valid for study purposes – this number is not clear for certain buildings are indicated as having been rebuilt and others not visited. On multi-building sites (such as Fore, Labbamolaga and Ardfert [Q 785 212]) it is not always evident which building was photographed. The present author has visited and examined the vast majority of the 149 sites.

accordance with available rock type. The majority of the churches which Ó Carragáin examined were constructed of sedimentary rocks. Most rocks of this type are deposited in water and are layered in beds which reflect differences in sediment supply or depositional rates, currents and composition. These differences are commonly very local and rarely constant. Individual beds may be from less than a millimetre to several metres in thickness, with the lateral thickness of an individual unit changing rapidly (Figures 5.10 and 5.11). The mason selecting building stone from an exposure, whether it was a natural cliff or a small quarry, would be confronted with these problems. Typically, the rock selected would have been taken from one bedded unit at a time, for rock normally splits far more readily along its bedding planes. If the bedded unit remains of approximately the same thickness over a desired wall length, a ‘course’ of masonry could have been created. Should the bedded unit vary in thickness, rise or dip would have resulted in the masonry course (Figure 5.12). In the work of Ó Carragáin (2005a) Plates III, V, X and XI showed this to varying degrees. Available stone block size must similarly depend greatly on bed thickness.

The nine distinguishing characteristics used in the 2005 analysis were: coursing, rise and dip of courses, stone block size range, whether certain stones were placed ‘on edge’, the stone shape, their fitting, the presence of ‘joggle joints’, the proportion of spalls (small, infilling and levelling fragments of rock), and the quantity of visible mortar. Unfortunately, all these features vary in

Subsequent to deposition many rocks will be altered by post-depositional changes. These may involve recrystallisation and additional cementation processes. The affects of such processes, together with differences in rock composition, may determine the shape of rock fragments and how well they can be fitted. Certain rocks splinter into irregular pieces; muddy limestones break

5.2 A brief review of the Ó Carragáin approach

155

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 5.5 The south wall of Clonkeen church, Limerick (R 689 547) viewed from the south-west clearly displays masonry of two distinct types (out of view, both terminate in an anta). The two rock types are of different lithologies and the masonry styles are of different ages: that at the lower western end has a number of ‘Patterned’ characteristics; that at the eastern end is possibly 15th century. Which of these two styles Ó’Carragáin (2005a) chose to analyse is uncertain.

Figure 5.6 The inside of the ruined early church at Kilree, Kilkenny (S 497 410) and the west side of the chancel arch reveal the use of at least three stone types and three or more periods of construction. 156

CHAPTER FIVE

Figure 5.7 The north wall of Clonamery church, Kilkenny (S 658 355), 2m. west of the infilled doorway viewed from the north (exterior). This displays worked granite blocks in courses supported by spalls. Compare with Figure 5.8.

Figure 5.8 The wall exactly opposing that shown in Figure 5.7 (that is, the inside of the north wall of Clonamery church 2m. west of the infilled doorway). This fabric is quite unlike that in the external surface, for instance it contains broken boulders and pieces of phyllite, etc. (Rule 325mm. long). 157

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 5.9 The south wall of Clonamery church, viewed from the inside (and directly opposite Figure 5.8) is constructed of a further variety of fabric, mainly granite boulders and phyllite, but in good courses. Compare with Figures 5.7 and 5.8. (Rule 325mm. long).

Figure 5.10 Lower Old Red Sandstones in a quarry in Mitcheldean, Forest of Dean (SO 671 185), illustrate the manner in which individual bands of rock may vary enormously in thickness. 158

CHAPTER FIVE

Figure 5.11 Cambrian slates and siltstones in cliffs at Fethard, Wexford (S 801 035) as in Figure 5.10, similarly show changes in thickness in individual beds.

Figure 5.12 The inside of the south wall of Killulta church, Limerick (R 434 534) viewed from the north-east (see section 4.5.6). The wall is made up of well-coursed blocks of Carboniferous limestone which have been worked from individual beds. One bed (or course) over the wall length of 4m. (approximately one-third up the wall) changes its thickness from 135mm. to 60mm.

into smaller pieces than pure limestones, etc. Joggle joints have been briefly discussed in section 4.5.19. It should be emphasized that because walls only provide a view in two dimensions it is extremely difficult to be assured that a true cut joggle joint is being observed. The mason working with massive large blocks will much more readily choose to remove a small portion of a block to permit it to fit to its neighbour than when building with small material (see also Harbison, 1972, 7). The variable quantity of mortar as compared with stone in masonry was referred to in section 4.5.25. That the use of cobbles or boulders of rounded shape would have increased the proportion of mortar was noted, but other factors can also influence the quantity. In some areas, the materials to make suitable mortar might have proved difficult to access; in certain instances (as in southern England) where these are in plentiful supply (and stone might be in short supply) the volume of a wall could be more than 50 per cent mortar). Finally, spalls, if available in quantity (a factor likely to be related to rock type), may influence the mason to use them to provide weak elements of pattern (Figure 5.13). However, they may occur in large numbers, where coursing is difficult, to assist in levelling, or to decrease the mortar requirements, etc.

The megalithic character of such stones has been observed and noted by others, such as, Champneys (1910, 29), Mytum (1982) and Harbison (1992, 20). Leask (1955, 51) further suggested, ‘…where wide and long but not necessarily very thick stones were obtainable … these were often set on edge, presenting to view a large surface of the natural bed. The appearance rather than the fact of great massiveness being thus attained’. Leask proposed that large stones were used in this way ‘in order to save mortar’. Leask’s proposal would not explain why during the building of ‘Patterned’ churches the stones were not used in the BVEB orientation. There would appear to be a simple explanation for the manner in which these large slab-like stones are used related to nothing more than their size; for placed BH or BVEB they would be too large to fit comfortably into the wall thickness. In a wall a metre thick, for example, any stone occupying much more than half of that thickness would create wall construction difficulties, only small stones being suitable for the other counterpart portion of the wall thickness. Many of the large slabs used are of such size that unless set BVFB they would project out of the opposite side of the wall (Figure 5.14). In many early church buildings megalithic stonework, when present, tends to occur mainly in the lower portions of the walls. This use must simply reflect that in the lower wall the larger stones will be more stable and that their insertion at this level would have required less lifting. More important perhaps from the point of view of the church historian, is that the mason building the church or wall must have had his supply of stones already available in order to select initially the larger stones. In the south of England for instance megalithic stones

Relatively thin beds, typically altered and strengthened by post-depositional changes, in some instances may permit slabs of large bedding plane area to be removed from a quarry without being broken. In many cases they may then have to be used ‘on edge’ (Figure 5.14). Such large slabs of rock certainly seem to have been selected where possible by the earliest masons. Indeed, their use extends well back into pre-Christian times: the preferred use always being for face bedded (BVFB) orientation. 159

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 5.13 The use of spalls (mainly of igneous rock) to semi-patterned effect in the south, largely granite and basalt, wall of boulders and irregular blocks, low in the chancel of the Maghera old ruined church, Derry (C 855 002). invariably have been extracted from a once, almost certainly local, Roman building. With experience and time, new techniques in working a particular rock type (possibly utilizing improved tools) might, no doubt, have materialized. Improved tools undoubtedly permitted the harder siliceous stone types to be more precisely broken (Potter, 1998), the cutting back of stone surfaces was almost certainly a late ‘Patterned’ period development (see section 2.9.3), and the ‘Romanesque’ masons much more readily created truly ashlar walls when suitable rock types were available. Wall thickness was a measurement that clearly varied over time in England (as shown by Taylor and Taylor, 1965), and in Ireland it again seems to be correct that generally ‘Romanesque’ walls are somewhat thicker. It is possible that some of the characteristics utilised by Ó Carragáin in his analysis, such as joggle joints or the use of spalls, might be employed differently or more gainfully in the more recent early masonry, but such changes have yet to be identified. In his final analysis, Ó Carragáin (2005a) identified five areas in which he believed there were similarities of masonry style. These were: ‘Northwest Clare/southwest Galway excluding the Aran Islands’, ‘The Aran Islands’, ‘Lough Corrib/Lough Mask’, ‘East Munster: East Cork, Mid and South Tipperary, Waterford and Tibberaghney, Co. Kilkenny’ and ‘Limerick, excluding Clonkeen and Ardpatrick’. Unfortunately, he then itemizes a variety of exceptions but fails to indicate precisely which churches are included in each group. Where the churches in West Cork or North Tipperary are classified, for instance, remains uncertain. For the remainder of Ireland, where

Figure 5.14 The north wall of the church at St John’s Point, Co. Down (J 528 338) viewed from the east end where the remnants of the north-east anta occur. One megalithic stone placed BVFR (face bedded in the north wall) is visible (rule 325mm.). The dimensions of this stone in this position make it obvious as to why it could not be placed horizontally (BH). 160

CHAPTER FIVE (p.129), ‘the geological diversity … was probably a more important factor’; styles appear to have varied.

In his further paper, Ó Carragáin (2005b) relies far less heavily on his masonry styles to suggest that regional variations in church building, largely during the 11th and 12th centuries, associated with environmental, economic and cultural differences, were possibly present in Ireland. On his other evidence supporting this view the present author offers no adverse criticism.

Only the Aran Islands of Ó Carragáin’s five areas can be readily geographically and geologically identified. All the churches on the islands are built upon, and of, rock from the same geological unit, Lower Carboniferous, Unit 64 of Table 4.1. According to the Geological Survey of Ireland (2006) these rocks were deposited in a shallow shelf sea, but even these accumulated as either fairly massively bedded limestones or thinly bedded calcareous mudstones. If both of these rock types were employed for building purposes, at least two masonry styles should be present plus any number of intermediary forms according to lithologies available. Because of the relatively bleak and unsheltered nature of the Islands the softer mudstones tend not to be exposed, instead the somewhat harder, purer limestones tend to be exposed and form the higher land and pavement areas (very much as also occurs in The Burren). This readily available stone was an obvious choice for early church builders, the rather irregular weathered blocks could be prised off the landscape much as was the practice by landscape gardeners until the very recent past. Size of the large building stone blocks might vary but little else, and small rock fragments (potential spalls) would be uncommon. That the blocks in the churches were removed from the pavement surface can be seen by the chemical weathering lines, which follow both bedding and jointing, observed on the block surfaces. Figures 4.93 to 4.113 all show these features to variable degrees in the masonry visible. As might be expected the masonry in the churches of The Burren (such as Temple Cronan, Figures 4.37 to 4.40) is very similar. A further example from The Burren, from Oughtmama (which church?), is selected to illustrate O Carragain’s north Clare/southwest Galway style (2005a, Plate II), as distinct from his Aran Islands masonry style (Plate V). Remembering that the limestone is slightly thinner bedded in the Oughtmama shot (it may be slightly muddier in content), the two masonry types are virtually identical. Put another way, if the two examples of Aran Islands style (Plates IV and V) are compared they are far more unlike than Plates II and V. Plate IV (of Temple MacDuagh) with an unusual number of spalls is atypical for the area or indeed the church (see Figures 4.112 and 4.113 and captions). Reference to section 4.5.34 in which Temple Macduagh is briefly described, indicates that Ó Carragáin’s Plate IV was actually a photograph of either of the north or south walls of the east end of the building. This walling is relatively modern and certainly not of his ‘preRomanesque’ or of the ‘Patterned’ style. The same style of masonry as illustrated in Plate IV, with ribbons of spalls placed vertically between large blocks as infill, may be observed in walls erected in the last few centuries. Unfortunately, the geological critique of the work described in Ó Carragáin (2005a) is that it is ‘speculative’ (p. 143), subjective, and lacking in scientific approach.

161

CHAPTER SIX THE ANTAE PUZZLE: A GEOLOGICAL INTERPRETATION FOR THE CONSTRUCTION OF ANTAE

Figure 6.1 The plan of an early, simple unicameral Irish church with antae is shown. Typically, there is a lack of apparent bonding between the gables and the adjoining walls and often this takes the form as illustrated. 6.1 Introduction Although the principal purpose of the intersecting finials, that are supposed to have been created by the extensions of the antae up the roof line of each gable, are generally considered to have been to support the ridge-piece to the roof, O’Keeffe’s case (1998) remains. A roof with a double-pitch can be far more simply created by supporting the ridge-piece on end rafters which rest in turn upon gable corbels. Corbels which must have served this purpose are evident on a number of early Irish churches such as the west church at Oughtmama (see section 4.5.10) and Temple Cronan (section 4.5.11) both in Clare, and Temple Chiaráin, on Inishmore (section 4.5.33).

The true nature and purpose of antae, constructed as they are at the corners of early Irish stone churches as prolongations of the north and south walls, has been a subject of some controversy for many years (Figure 6.1). Although antae resemble buttresses, they are not built against the walls which carry the greatest lateral roof pressures, and this possible function is universally discredited (Leask, 1955, 55; Hare and Hamlin, 1986, 143). Their similarity to the projecting walls used to support the megaron porch recorded in early Eastern Mediterranean cities such as Troy and Mycenae promoted the suggestion (Champneys, 1910, 38) that they were modelled on both these and later similar structures. Others (as Petrie, 1845, 186), have proposed that they may be a specifically Irish version of the ornamental English and continental pilaster or lesene. Probably, the most commonly held view is that described by Leask (1955, 55-6); that they were constructed to support the roofing timbers and that they were a skeuomorphic derivation from elbow-shaped cruck timbers that had served this purpose in earlier wooden churches. This view has recently been questioned. O’Keeffe (1998) observed that the building of such large structures, ‘simply to carry the barge-boards of a roof when projecting corbels at eaves level could fulfil the duty as easily’, was difficult to substantiate.

A further view that antae, having been built no higher than the eaves of the gables, served the same role as corbels to support the end rafters (see section 4.5.19), no more successfully counters the arguments offered above. 6.2 The evidence 6.2.1 Evidence for possible roof support There appears to be no known example of a church with antae also displaying gable corbels to support the church roof structure. With antae protectively continuing up the full height of the quoins, at gable and then roof height, 162

CHAPTER SIX they certainly could readily provide support for the ends of the wall plates, principal rafters and possible tie-beams of the roof, as proposed by Leask (1955). However, O’Keeffe (2003, 70) strengthened his argument against this interpretation when he wrote that antae usually rise ‘…no further tha(n) the point at which the ascending walls of the east and west ends of the church me(e)t the level of the wall-head along the sides’. There are probably just two exceptions (discussed below) where the antae are typically cited as rising to the ridge piece. Instances, more generally, may be observed to remain only to the head of the north and south walls, as suggested by O’Keeffe. For example, this occurs at Glendalough Cathedral (see section 4.5.19 and Figure 4.66), where the antae seem to have been capped with horizontally bedded stones to provide support; much as would have been afforded by corbels (at Glendalough, and some other churches the capping stones are probably relatively modern). Stone roofs, supported by wood, are relatively impermanent structures and in no churches with antae do the original roofs remain. Leask (1955, 43) stated that there was no evidence for the expedient of tiebeams in early stone roof structures. He argued the case for the earliest Irish wooden roof structures being built on the cruck principle and the subsequent skeuomorphic development of antae being applied to stone roofs. (Two minor arguments against this suggestion could also be made. Large oak trees, typically used for creating crucks, are uncommon throughout much of Ireland. In addition, in longer buildings crucks were inserted at intervals, but there is no evidence of these in Ireland). Certainly, by extending the antae up the gable ends to terminate in intersecting finials the ends of roof purlins and the ridgepiece could be further supported, although they could have been equally carried indirectly onto the gable wall.

Figure 6.2 Kilmalkedar church, Dingle, Kerry (Q 403 062) is viewed here to exhibit its west gable. The antae rise to roof height where they are ‘capped’ with slightly broader stones; they then continue a very short distance in line with the roof. The finials are, however, constructed on top of the gable and are not associated with the antae. There are at least two, and probably three building periods involved in the construction of the west gable wall, the earliest in the wall rising only to the height of the top of the arch jambs.

The island church of St MacDara, described by Harbison (1992, 163) as possessing antae that ‘continue right up the gable and meet at the top’, is now rebuilt at the higher gable levels to the antae pattern proposed by Leask. A picture of the church (Harbison, 1982, Plate 5b) shows that in its reconstructed form the ridge-piece could not have been supported on the finials in the manner as proposed by Leask (1955, 44). The early Irish roof may not have utilised purlins: were they used; there is no indication of cavities for their support in the stonework of the antae extensions shown in the early engraving of St MacDara (figured in Petrie, 1845, 187). A wooden ridgepiece would also have been difficult to insert into the roof that was then visible.

beams, themselves supporting crosses and other iconographic forms, were suspended on the tops of the antae’. The evidence, from actual examples, for antae being used wholly for roof support, is, therefore, virtually nonexistent. Finials displayed on church-like features on high crosses, such as the ‘cap’ stone on Muiredach’s Cross, Monasterboice (O 043 821), noticeably are constructed on gable ends not antae extensions (Figure 6.3). Certain early stone tombs, as that seen in the graveyard at Clones round tower (H 502 256), have been similarly offered as representing reproductions of the finials of church antae (Figure 6.4). Again, these finials are placed on the gable ends of the tombs, and Harbison (1999) believed the tombs to be copies of original metal shrines rather than churches.

The second church that is cited as typically exhibiting evidence of gable extensions to antae, is Kilmalkedar, in Kerry (Q 403 062). In the case of this church, although it carries finials they are built on to the top of the gable walls and not the gable antae extensions (Figure 6.2). On the limited evidence available, the purpose of finials would appear to more probably have been iconographic (as proposed by O’Keeffe, 1998, 119) rather than for the provision of structural support for the church roof. In 2003, therefore, O’Keeffe (p. 70) proposed that one ‘…might consider instead the possibility that horizontal

Following a recent examination of the stonework of the more accessible of the early Irish churches which preserve antae or remnants of these structures, the present author now suggests that one very significant reason for 163

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 6.4 Traditionally described as the shrine of St Tighernach, this tomb in the churchyard of Clones church, Monaghan (H 502 256) is constructed with finials. It has been suggested that structures such as this support the contention that they were modelled on existing churches, although Harbison (1999) offered a different view. The tomb possesses no antae and the finials are built on the top of the gable.

Figure 6.3 Muiredach’s Cross, Monasterboice (O 043 821), at its head carries a ‘church-like’ structure bearing finials. These finials are noticeably on the top of the gables. the construction of antae has been overlooked. The orientations of the bedding of the stonework of antae reveal when the antae were built or repaired. Additionally, the structural relationships of the antae walls with their adjoining gable walls is remarkable and requires explanation.

independently of the walls) as supports, they not being bonded into each other. Every second course of the gables has, however, a stone slightly inserted into the side walls.’ Biggar (1896) The present author recently examined the more accessible Irish churches with antae (Figure 6.5). Unaware of Biggar’s findings, he too noted that the gable walls appear initially, in many instances, to lack any proper stone bonding with their respective north and south walls. One wall appears to abut directly upon the other. Exceptions tend to occur only where walls have been repaired. In the majority of instances, on the exterior of each church the abutting walls are those containing the ‘west’ doorway and the ‘east’ window, that is, the gable walls. The external appearance is, therefore, of the north and south antae walls having been built first (Figures 6.1, 6.6 and 6.7). The straight junctions between the walls obviously appear to suggest lines of weakness in the church structure. As noted by Biggar, detailed scrutiny of the junction shows that slight bonding between the walls does actually exist, with stones perhaps every three or four courses inserted by no more than a few centimetres into the opposing wall (Figures 6.8 to 6.12).

6.2.2 The structural evidence The occurrence of antae in early Irish churches appears to be confined to the walls adjoining the corners of the gable ends of a limited number of simple, single-celled buildings, or buildings which originally were constructed to this design. The Christian origins of these stone buildings result in these gables typically facing to the east and the west. One of the earliest detailed descriptions of a church with antae was provided by Biggar (1896) following a visit to St MacDara on the Atlantic island of the same name, off Galway. Part of his description is worthy of repetition: ‘The north and south side walls project about a foot (305 mm.) beyond the east and west gables, and bear the roof directly upon them, apparently using the gables (which have the appearance of being built 164

CHAPTER SIX

Figure 6.5 A map of Ireland to illustrate the position of those churches which are known to possess antae or traces of antae. Those churches detailed in Table 6.1 are illustrated with a solid circle and those not examined are shown with a cross. A few churches for which traces of antae have been claimed, such as Ardpatrick (R 637 208), are omitted because no evidence of an anta could be seen. All church sites are close to navigable waters. Area A is drained by the River Shannon and its tributaries; in Area B the churches may be reached by the river systems of the Suir, Nore and Barrow draining into the Waterford Estuary; and in Area C the River Blackwater flows into the Youghal Estuary.

165

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 6.6 The lower portion of the north-west anta at Glendalough Cathedral. Note that the anta wall appears to have been built prior to the gable wall (to the right) and the two walls do not appear to be bonded.

Figure 6.7 The shallow, south-west anta of Kilmalkedar church, Dingle, appears to lack bonding between the anta wall and the west gable wall and like Figure 6.6 suggests a line of weakness in the construction.

Figure 6.8 At Clara church, Kilkenny (S 578 523) the south-east anta wall is slightly displaced from the east gable wall. Bonding between the two walls when viewed from the east appears to be negligible and of the two walls in this view the gable wall seems to be of a more recent date. 166

CHAPTER SIX

Figure 6.9 In the last few years the north-east anta at Sheastown church, Kilkenny (S 544 523) has started to fall away from its adjoining east gable wall. The east end of the site is viewed in this photograph. Figures 6.10 to 6.12 display the relationship in more detail.

Figure 6.11 Viewed from the south the same anta as in Figures 6.9 and 6.10 reveals the limited bonding between the gable and anta walls. Holes created to accommodate two gable wall stones can be observed in the anta wall.

Figure 6.10 A closer view of the north-east anta of Sheastown church.

167

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 6.12 The same Sheastown north-east anta is now viewed from the north west and much of the anta wall can be seen to have already fallen leaving it unsupported. Inside the church, in most instances, where the original walls remain, a similar arrangement exists; the walls containing the antae having apparently again been built first. However, the arrangement of the abutment may sometimes be reversed. These remarkable structural relationships certainly seemed to require explanation (see section 6.4.1).

where the structure was partly of timber (Waterman,1967), and in Co. Longford, at the church of St Mel, Ardagh (Harbison, 1982, 628). Some stone church buildings are certainly thought to have existed, however, before the 9th century. Hamlin (1984) has proposed that a stone building existed at Duleek, in Meath (see section 4.5.24), in the 7th century. Berger (1992) radiocarbon dated early mortar from a range of churches to the 7th and 8th centuries. It would seem possible that, prior to the need for a defended building, some stone churches may have been built along a style similar to that of early Anglo-Saxon England, that is in the ‘Patterned’ style, making some use of vertically bedded stones for decorative purposes. Certainly, there is ample evidence of the use of vertically bedded stones in the walls of many early monastic sites both in Ireland, and in Scotland where the sites at Capel Finian (NX 279 489) and Barhobble (NX 311 494), both in Dumfries and Galloway, may be cited as examples. In England, the quoin stones of Bradwell-on-Sea (TM 031 082) in Essex, a church probably built in the mid-7th century (Taylor and Taylor, 1965) provide an early example of vertical bedding orientation.

6.3 The physical protection of early Irish churches 6.3.1 What was the need? Evidence from the various Annals and other hagiographical literature indicates that church properties and settlements in Ireland first suffered the impacts of Viking and other raids about the commencement of the 9th century. Laing (1977, 225) recorded the first raid by the Norse as occurring in 795. The recent analysis of the Annals, by Manning (2000), revealed that churches built of wood were especially susceptible to destruction by burning. Furthermore, even a number of daimhlaig (damliac) buildings, interpreted as stone churches, such as at Kells (in 920 and 1060) and Kilfenora (in 1055) were destroyed or burnt. Authors (such as Leask, 1955, 2; Harbison, 1992, 15) agree that these raids must have given an impetus to the construction of stone, in place of wooden, churches (see section 4.1). As evidence of this process, instances of stone churches with antae being preceded by earlier timber structures have been described following excavations in places like Co. Down, at Derry,

It seems probable that in the early part of the 9th century, Irish churches, whether of wood or stone, had to try to offer some defensive security, for valuables, if not for persons. The stone buildings appear to have been small single-celled rectangular structures. These buildings were 168

CHAPTER SIX

Figure 6.13 Megalithic face-bedded stones built into the lower portion of the east gable wall at St Mel, Ardagh, Longford (N 204686). Constructed between the antae these stones would probably have assisted in the defence of the church. typically provided with relatively high walls, a small comparatively inaccessible east window, and a thick wooden barricaded west door (see section 7.1.7). The strength and resistance of the building and its walls would have been an important consideration.

The random, rather than patterned, use of face-bedded large stones, particularly low in walls, cannot have been for ornamental reasons. Section 5.2 describes the probable reason for their common occurrence. The flat bedding face of such stones tends to provide a smooth surface finish to a wall. Additionally, as implied by Leask (1955, 51), the appearance of ‘great massiveness’ must have portrayed both strength and solidity. It seems probable that this applied in both a religious and a physical sense, and these megalithic stones used randomly and low in the walls, offered a level of material protection against damage (Figure 6.13).

6.3.2 Megalithic random stonework When the stonework of early monastic and church sites in Ireland is reviewed, the megalithic character of many of the stones, particularly in the lower portions of walls, is very apparent (for discussion, see section 5.2). Frequently, the same large stones are placed in the wall with their bedding orientated vertically. The emplacement of wall stones which have their natural bedding or stratification in a vertical attitude has been shown to be of very great significance in early Christian ecclesiastical buildings. Its occurrence is widespread throughout Ireland, where it may be observed in sites as dispersed, in kind and geography, as Reask monastic site in Kerry (Q 367 045), St Mel, Ardagh, Longford (N 204 686), or, where the rock type is a lineated granite, as at the oratory of St James, St Mullins, Carlow (S 729 329). In key structural features such as quoins, antae, and the jambs of arches, vertically orientated bedding in emplaced stones has proved critical to the identification of Irish ‘Patterned’ workmanship (see section 2.7.3). There is no doubt that the early builders, throughout the British Isles, had a penchant for vertically bedded stones, and that their use in wall faces was also undoubtedly at least in part for decorative effect (see section 4.9).

Megalithic stonework in the wall faces would have provided, at least in appearance, an increased level of protection (and even the smooth face of any face bedded stones could have deterred the climbing of the walls by potential assailants). In these circumstances the corners of a rectangular building would have offered points of least resistance (Figure 6.14B). This would have been especially the case where quoin stones were set in the typical alternating, vertically bedded style of the period (Potter, 2005b; 2006d). Individual stones could have been readily prised off from the quoin to expose the weaker wall core. However, the church corners could be protected. It is suggested that the development of antae may have been an ingenious device for this purpose, for the addition of the supplementary third wall would provide added strength and protection to the corners.

169

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 6.14 A. A simplified sketch-section in the horizontal plane of a possible arrangement of interlocking stones in the wall behind a north-west anta is shown. Mortar is shown in black and individual interlocking stones are shaded differently. With the exception of the facing stones other stones cannot be removed without complete destruction of the walls. B. To a smaller scale, a similar section is shown of a typical side-alternate quoin. Once the large external stones have been removed others behind them (which would have been difficult to fully interlock) could also be separated from the walls. 6.3.3 Protecting a small church building – the use of antae

with antae, such as Loorha (see section 4.5.3), still retain some evidence of the strong bonded nature of the antae corners.

In the construction of antae, the four points of weakness to the single-celled church, the quoins, are no longer present (Figure 6.14A). Each corner would be greatly strengthened and yet ornamented stonework was still feasible. The stonework in the earliest, ‘Patterned’ style, antae still adopted the early practice of placing many stones with their bedding or lineation orientated vertically, and an element of pattern was often illustrated (Figures 6.6 and 6.15). Potential assailants, observing that the antae were no more than extensions of the north and south walls, would leave these structures free from wilful damage, for their destruction would not provide access to the church interior. In building the walls behind each anta it would have been possible to interlock the stones within the wall so that they could not be prised or levered away from the wall (Figure 6.14A). At the church corners the intersection of stones, with the inclusion of the antae, from three wall directions (rather than two as occurs at a quoin) would have provided also an excellent facility for interlocking the stones. The whole might then be regarded as somewhat similar to the interlocking ‘Chinese puzzle’. Only withdrawal by lifting stones vertically would remove any linkage between the stones. If this technique was applied, the interlocking of stones would probably be necessary only at wall levels which were particularly susceptible to attack. Certain churches

6.3.4 The distribution of churches with antae In 1982, Harbison classified early Irish stone churches according to their morphology into various types (see section 4.1). Those churches with antae were regarded as being of sufficient distinction and adequate abundance to be elected to a particular type. More recently, Ó Carragáin (2005a; 2005b) examined the distribution of the various church types in Ireland (see Chapter 5), to conclude that of the five church types which he chose to elect, some, such as the small corbelled oratories were regional in distribution. He had much greater difficulties in attempting to account for the distribution of the churches with antae (2005b, 33). For those in east Munster for instance, he wrote, ‘…the reasons why the type was perpetuated there remain obscure’. It is, therefore, proposed that the original primary purpose for the construction of antae was to provide an ingenious defensive mechanism to protect the relatively defensively weak corners of early single-celled Christian stone churches. The distribution of churches with antae (Figure 6.5) does much to support this view. The vast majority of the churches preserving antae are to be found in close proximity to the sea or navigable rivers like the Shannon 170

CHAPTER SIX

Figure 6.15 The north-west anta of Temple Bhreacháin, Inishmore (L 811 121). The anta wall gives the impression of having been built first and being only poorly bonded to the gable wall. The anta is constructed of stones which are ornamentally inserted into the structure in ‘Patterned’ fashion. Note that some of the stones are face-bedded whilst others are edge-bedded (see section 4.5.32 for detail).

Figure 6.16 The north-east anta to the early 12th century nave of Dungiven ‘priory’, Derry (C 694 084) has been ornamented with ‘Romanesque’ clasping corners. The rule is 325mm. long. Note that in typical ‘Norman’ style of England, the corner stones have been cut back.

and Suir/Nore/Barrow system, situations that were particularly susceptible to Viking or other similar attack. Of the churches which today display evidence of antae, none is as much as 10 km. from waters which would have been readily navigable. Ardagh, approximately 8 km. from the River Inny, a tributary of the Shannon, is probably the most distant. The distribution of churches with antae is not unlike that of the earlier Irish round towers; structures, also thought to have been built, if primarily for bell-ringing, at least in part to provide some protection for valuables or sanctuary from marauders (Hare and Hamlin, 1986, 143; Berger, 1992, 888; O’Keeffe, 2003, 74).

end of the nave are decorated with early 12th century ‘Romanesque’ clasping quoins (Figure 6.16). 6.4 Other aspects of antae construction 6.4.1 Apparent frailty related to lack of wall bonding This study, in suggesting that antae were erected as an ingenious means of protection for the vulnerable corners of small early stone churches, is no more unrealistic than the various theses presented by Leask (1955) and others. Stonemasons, in particular, recognise the wisdom in adding a third wall to vulnerable corners. On first impression, however, it suffers from an apparent building frailty; the observed lack of bonding between the gable and antae walls (Figure 6.1). Initially, the present author suspected that what was an apparent lack of wall bonding was provided intentionally by the early Christian mason as an act of deception to frustrate the assailant, guiding that person to an area in the wall from which it would

Antae continued to be used apparently into ‘Romanesque’ times and until the risk of attack from potential assailants had diminished. ‘Romanesque’ BH placed stonework can be identified in the antae of a number of churches, as for example, Britway, Cork (W 882 882). In Ulster, at Dungiven ‘priory’ (C 694 084) antae preserved at the east

171

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND prove impossible to remove further stones. Those persons intent on destruction or entry, would have observed the apparent lack of bonding between the gables and the north and south walls, and under the notion that these marked errors in construction and points of weakness, they might attempt to destroy that part of the building that was probably the strongest.

would appear to have been erected in the ‘Romanesque’ period. Once built, a minor subsidiary advantage of the antae results: they would of course have provided some weather protection, much in the manner of a roofless porch. If this function is accepted, a case would be provided against the shortening of antae over time. As a general observation it would perhaps be more appropriate to argue that the largest antae, in both depth and width, relate to the largest church buildings.

Such early Christian guile concerned the author, so much so, that those churches with antae were revisited. Wherever possible, if others were present they were invited to judge ‘Which wall was built first?’ Not that one wall could have been built before the other; rather the walls were built to look that way. Certainly, only limited wall bonding exists in many instances. Such astuteness remains possible, but examination of the bonding between walls of different ages, as between nave and later chancel walls, often takes a similar form. On balance, the suspicion must be that the masons of the period chose to provide only limited bonding for adjoining walls, finding that this met the necessary requirements for the strength of their buildings.

Although the widths of most antae tend to relate approximately to the widths of the related north or south antae walls, Glendalough Cathedral is not the only instance where the walls and the antae widths differ (see section 4.5.19). The church at Kiltiernan (see section 4.5.38), cited in this work, provides a further example (Figure 4.122).

6.4.2 Depth and size of antae Manning (1998, 76), Ó Carragáin (2005a, 139), and others, have suggested that the amount which antae project beyond their respective gable walls (their depth) might reflect the date of origin of these structures. They proposed that those of earliest origin were the examples which projected the greatest. Ó Carragáin cited, for instance, the church at Clonkeen, County Limerick (R 689 547), where he stated, ‘…the pre-Romanesque antae …are 0.66m deep; those of its Romanesque extension are 0.30m deep’. Many of the antae examined have been extensively rebuilt, and Clonkeen provides no exception. All of the antae at Clonkeen have been rebuilt (Figure 6.17), with some of the rebuilding including stones dressed and inserted in recent centuries. Labbamolaga small church (see section 4.5.14 and Figures 4.46 and 4.47) provides a typical example of the problems of rebuilt antae and the difficulties of trying to assess their date of construction from the differences in amounts the antae project from their associated gable walls. In view of the variation observed in the antae depth of the churches examined, the present author recorded the depth and width of the majority of these structures in the churches visited (Table 6.1, in this Table the antae are measured to the nearest centimetre). Of the ‘Patterned’ antae listed in Table 6.1, only nine can be adjudged to be sufficiently preserved to remain, at least in part, of original depth. The nine include St Bhreacáin (anta depth 170mm.) and Temple Macduagh (330mm.), both on Inishmore, with the smallest projection; and Glendalough Cathedral (600, 640, 680, 630mm.) and Loorha (650mm.), with the largest. Others in the stated nine are: Agha, Ardagh, Fore, Killeenemeer and Tuamgraney. With so many antae showing evidence of reconstruction, it is somewhat more difficult to confirm the age of the ‘post-Patterned’ forms – although Britway (300mm.), Kilree (310 to 340mm.) and Leighmore (280 to 310mm.)

Figure 6.17 The west face of Clonkeen church (R 689 547) as viewed from the south-west. Although the lowest part of the west gable wall (and short stretches of the north and south walls) are probably of ‘Patterned’ age, both the west antae have been entirely rebuilt (probably, in part, since the church fell into disuse in the mid-17th century), with some of the stones being relatively modern. The stonework settings of the west doorway are ‘Romanesque’ (possibly partly including re-used earlier stones). 172

CHAPTER SIX Table 6.1 A table illustrating the variable lengths which antae project from their respective gable walls. The breadth of each anta is also given together with the likely age of their construction as determined from their stonework Area A. Churches with antae, in the Shannon valley. Church site and County

Grid Reference

Antae Dimensions1 NW SW NE SE d. w. d. w. d. w. d. w.

Building Style and Detail2

Ardagh, Longford

N204686

51 82 51 81 58 89 48 85

Clonfert, Galway

M962212

64 103 68 103 ? ? ? ?

Clonkeen, Limerick Clonmacnoise, Offaly (Cathedral) Clonmacnoise, Offaly (T. Ciarán) Clonmacnoise, Offaly(T.Dowling) Fore, Westmeath (St Feichin) Loorha, North Tipperary Lynally, Offaly Tuamgraney, Clare

R689547

Lowest ‘Patterned’,rest rebuilt Rebuilt, ‘Romanesque’ evidence only Rebuilt, especially east end where medieval Rebuilt, ‘Patterned’ traces; NW ‘Romanesque’ All rebuilt, NE with ‘Patterned’ traces Rebuilt, ‘Patterned’ traces

Not

determined

N010306

70 85 71 90 74 91 71 89

N010306

Not determined NP

N010306

22 69 20 68

? ? ? ?

N510704

59 87 54 85

? 86 ? 85

M920046

65 105 65 105 ? 104 ? 107

N297239

36 85 37 85

NP

NP

R637830

57 86 61 85

NP

NP

NP

Lowest ‘Patterned’, rest rebuilt Lowest ‘Patterned’, rest rebuilt Insufficient visible to determine Lowest ‘Patterned’, rest rebuilt

Rock Type3

Carboniferous calcareous sst Carboniferous limestone Mixed Mixed, fluvioglacial Mixed, fluvioglacial Mixed, fluvioglacial Carboniferous limestone Carboniferous limestone Carboniferous limestone Upper Devonian sandstone

Area B. Churches with antae, in the Suir, Nore, Barrow valleys. Agha, Carlow Clara, Kilkenny Clonamery, Kilkenny Killeshin, Laois Killoughternane, Carlow Kilree, Kilkenny Leighmore, North Tipperary (small) Moone, Kildare Rosecrea, North Tipperary Sheastown, Kilkenny Tullaherin, Kilkenny

S730654 S578564

?

? NP

54 c70 NP

NP

NP

? 74 27 74

S658354

26 78 35 81

NP

NP

S673778

23 97

NP

NP

S777542

22 78 24 78 27 77 23 75

S497410

34 83 31 86 33 87 34 ?

S225576

31 75 30 78 29 67 28 76

N790927 S138895

NP

NP

NP

58 88 58 88

? 71 54 81 NP

NP

S544523

NP

NP

28 77 25 80

S591478

NP

NP

? ?

45 84

173

Lowest ‘Patterned’, rest rebuilt No evidence of ‘Patterned’ Not determined, probably ‘Patterned’ origin Rebuilt, no evidence of ‘Patterned’ No evidence, probably ‘Romanesque’ ‘Romanesque’ ‘Romanesque’, some rebuild Rebuilt, no evidence Rebuilt, ‘Romanesque’but evidence of ‘Patterned’ Rebuilt, ‘Romanesque’, low NE trace ‘Patterned’ Little clear evidence

Granite Carboniferous limestone Granite Granite (mainly) Granite Carboniferous limestone Carboniferous limestone Silurian greywacke Upper Devonian sandstone Carboniferous limestone Carboniferous

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND Area C. Churches with antae, in the Blackwater and tributaries valleys. Britway, Co. Cork Coole, Co. Cork (large) Coole, Co. Cork (small) Killeenemeer, Co. Cork Labbamolag(g)a, Co. Cork (small)

W882882

30 76 30 74

NP

NP

‘Romanesque’

W860952

39 71

NP

NP

NP

W860950

NP

NP

Evidence limited, ‘Romanesque’ traces Rebuilt, ‘Patterned’ traces

25 77 28 87

R775070

26 78 27 81

‘NP’

‘NP’

R764176

42 70 45 70 57 72 64 72

‘Patterned’, rebuilt at top Rebuilt, ‘Patterned’ and ‘Romanesque’ traces

Upper Devonian sandstone Upper Devonian sandstone Mixed, original Devonian sst Upper Devonian sandstone Upper Devonian sandstone

Area D. Churches with antae, in coastal margins. Ardmore, St Declan, Waterford Derry, Co. Down (small) Dulane, Meath Dungiven, Co. Derry Glendalough, C’dral, Wicklow Glendalough, Gateway, Wicklow Inishmore, Galway St Bhreacáin Inishmore, Galway Temple Macduagh Kill of the Grange, Dublin Kilmalkedar, Kerry Kiltiernan, Galway Nendrum, Co. Down Ratass, Kerry St John’s Point, Co. Down

X189774

43 58 48 62 46 59 46 59

Rebuilt, ‘Patterned’ traces

J613523

45 67 42 75 45 80 41 c80

N741788

74 82 73 84

Rebuilt, possibly ‘Romanesque’ Rebuilt, evidence limited, ‘Romanesque’ traces Rebuilt, ‘Romanesque’ evidence (Figure 6.16) ‘Patterned’

C694085

NP

NP

NP

NP

? 53 ?

51

T123968

60 82 64 85 68 82 63 81

T124969

37 ?

37 ?

L811121

17 82

NP

L823104

33 76 33 ?

O226273

25 83 25 82 28 83 31 83

Q403062

Not

40 85 39 82 NP

NP

? ?

? ?

26 81 28 82 ? 89 ?

J524637

71 77 71 75 NP

Q853141 J528338

Not

‘Patterned’

determined

M437156

NP

determined

37 66 38 66 NP

NP

Rebuilt, some ‘Romanesque’ evidence Top rebuilt, ‘Patterned’

Rebuilt, no evidence Rebuilt, probably recent

?

Rebuilt, ‘Patterned’ traces

Upper Devonian sandstone Silurian greywacke Carboniferous limestone Carboniferous sandstone Ordovician phyllite/schist Ordovician phyllite/schist Carboniferous limestone Carboniferous limestone Mixed, originally granite Mixed

Rebuilt, no evidence

Carboniferous limestone Quartz dolerite

Rebuilt, now ‘Romanesque’ Limited evidence, rebuilt

Upper Devonian sandstones Mixed

Notes: 1 In this column the following abbreviations are used: NW, etc. = points of compass d = depth of anta measurements in centimetres w = width of anta measurements in centimetres ? = unable to determine or not measured NP = not present Measurements given are in centimetres and recorded to the nearest centimetre (i.e. the nearest ten millimetres). 2

Details in this column refer only to the antae, the building might show supplementary or different evidence.

3

sst = sandstone Mixed = only used where extensive rebuilding has involved many replacement and different stone lithologies.

174

CHAPTER SEVEN RESERVATIONS AND CONCLUSIONS

number of occasions, and that few Irish ecclesiastical walls represent work of only one period. Unfortunately, Berger failed to present the precise details of the locations of his samples, for at the time great swathes of most walls were regarded as being of the same age. Greater precision in the identity of these localities might well provide more acceptable dates for his samples and greater approval for his analyses.

7.1 Reservations and outstanding problems 7.1.1 Introduction In many respects this work is incomplete, for ecclesiastical geology is still in its infancy. In time, no doubt, further aspects of both stone use and masonry construction for different periods of workmanship will be noted. Features of ‘Romanesque’ work have, as yet, been barely examined and various later styles virtually completely neglected. The long Roman (or equivalent Celtic) to ‘Romanesque’ time interval is currently united under the term ‘Patterned’, or in England ‘Anglo-Saxon’. Changes in technique, such as the introduction of doorway architraves and cut backs, have transpired during this long interval. Currently, the dates when these may have occurred are only approximate and lacking any precision. Nor indeed, does knowledge exist as to what may have motivated these and similar changes. It should be noted that in using ‘Celtic’, or Celtic Christian, to describe the earliest church in Ireland or Scotland, Thomas (1971, 5) made the point that there ‘never was a “Celtic Church”’ for, ‘The official language of the church in the Celtic-speaking parts of the British Isles was Latin’. This observation, together with their respective geographical situations, raises the question as to the extent of communication between the Romans and Christian Celts and, therefore, the levels of transmission of ideas that might have existed in the fields of very early ecclesiastical construction.

7.1.3 Were the external surfaces of ‘Patterned’ ecclesiastical buildings limewashed or plastered? The present author has written extensively on this subject (Potter, 2006c; 2008d), for if external surfaces were covered, the whole purpose of decorated stonework on external church surfaces would be defeated, and to the mason the work would have been deemed unnecessary. Although there is ample indication that plaster or limewash was frequently applied during the ‘Romanesque’ period and its application probably resulted in the rapid transition from ‘Patterned’ to the ‘Romanesque’ fashion (see section 2.8), no evidence of Anglo-Saxon or ‘Patterned’ use of such covers can be discovered. All English supposed incidences of external plasters on ‘Patterned’ period buildings have been investigated and found to be erroneous. The record by Leask (1930, 130) which was evident at the time of demolition of the Friar’s Island church (see section 4.5.8), therefore, requires some examination. Unfortunately, the south-east face on which the plaster was preserved is now invisible as a result of the rebuilding. Examples of post ‘Patterned’ plaster coating of earlier ‘Patterned’ external walls exist (Potter, 2008d) and the likelihood is that Friar’s Island provides a further instance, but without visible re-examination confirmation is impossible (see also section 7.2.2).

A wide range of features related to stone use remain available for study. Some of these have been referred to within the general text of this work. Others, such as the forms and varieties of windows over the ‘Patterned’/ ‘Romanesque’ interval, are likely to reveal an enormity of information on fashion changes over the period. In the general context of this work, a small number of matters related to early church and ecclesiastical construction remain and some of these can now be briefly elaborated upon.

7.1.4 Areas of potential analysis - plinths To assist in the determination of the different periods of church building, several aspects of early church construction would benefit from much more detailed analysis. In England, Taylor and Taylor (1965, 13) indicated that simple square-section wall plinths, particularly when constructed of large flat stones were ‘a good indication of Anglo-Saxon workmanship’. Plinths of this type exist beneath the walls of many of the churches listed in Table 4.2. Examples may be seen at Kilmacduagh Cathedral (beneath west wall); Kiltiernan (early part of church), and Oughtmama West church. They also occur as footings in various other churches, some of which are represented by walls of only restricted height. In this latter category the oratory at Annaghdown, Galway (M 287 381), in which the walls are never more than three courses high, may be cited. Each of the four

7.1.2 Analysis of mortars When Berger (1992; 1995) provided radiocarbon dates for the early mortars of a number of Irish ecclesiastical buildings of early uncertain age, the answers to unravelling the complexities of their age relationships appeared to be near. Furthermore, it seemed possible that the same technique could be applied more widely. A number of church historians, however, criticized the work, at least verbally, for its inaccuracies – the dates were unreliable and often, in their belief, incorrect. It is now very evident, hopefully from the present studies that the masonry of many early churches has been rebuilt on a 175

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 7.1 Viewed from the north-west, Oughtmama East church, Clare (M 305 079) displays a typical early style plinth. The stones in the quoin above the plinth rise as BH, BH, BVFL, BVFR, but the two highest stones are probably not original. More study of the plinth might confirm that this church was originally of ‘Patterned’ origin. The rule is 325mm. long. quoins in the oratory contains a stone that is vertically bedded, but which could have been emplaced in this way during modern reconstruction. The plinth, however, tends to support their pre Anglo-Norman originality. The case related to Oughtmama East church, Clare (M 305 079) is similar (Figure 7.1). There, only the two lowest stones in each quoin (perhaps five in the NE quoin), and where present, the plinth, might be original.

free-standing, unaltered ‘Patterned’ walls than there are in England. Furthermore, with many early ‘Patterned’ churches being extended in length in ‘Romanesque’ or more recent times, there has been a tendency for the walls, with extension, to be continued at the original width.

In Ireland, as in Scotland, many churches, as the last two aforementioned examples, remain only as ruins rising but a few stone courses. In these circumstances, use of stone bedding orientation in structures like quoins, as detailed in this work, then becomes less reliable, because of the insufficient wall height and the limited number of quoin stones. Indeed, in many instances the superior and often better shaped stones of the corners of buildings no longer remain, for they have been typically the first to be removed for use elsewhere.

Although this work has briefly examined Scotland’s last two free-standing round towers, at Abernethy (see section 3.3.4) and Brechin (3.3.5), to illustrate that both reveal distinctive characteristics of the ‘Patterned’ period: no attempt has been made to provide details regarding the far more numerous comparable Irish towers. Such an examination would present a task in its own right, largely because their shape in itself presents major difficulties. The analysis of masonry requires detailed scrutiny generally involving the use of a hand lens and is, for this reason, most commonly restricted to only the lowest levels of a building. For higher levels, some assistance can be gained with the use of high-powered binoculars and the telephoto facilities of a digital camera, but adequate and suitable lighting then becomes particularly critical.

7.1.6 What of Irish Round Towers?

7.1.5 Changes in wall thickness Taylor and Taylor (1965, 12) and Taylor (1978, 959) demonstrated that in England Anglo-Saxon walls were generally thinner than those walls erected during the Norman period. It seems likely that the same broad principle of ‘Patterned’ walls being thinner than ‘Romanesque’ walls applied also in Scotland and Ireland. In both these countries there are unfortunately far fewer

A very wide range of authors have contributed to the existing knowledge concerning Irish round towers: some of the most significant of these being authors like Petrie 176

CHAPTER SEVEN

Figure 7.2 The round tower at Glendalough (T 123 968) is here viewed from the south-east. (1845); Barrow (1979); Hamlin (in Hare and Hamlin, 1986); Lalor (1999); Stalley (2000) and O’Keeffe (2004). The last of these authors listed 97 certain and alleged round tower sites in Ireland and indicated that as many as 64 (p. 43) might, at least in part, still be visible. The present author has visited just 30 of these. O’Keeffe (2004) outlines much of the current knowledge with regard to the towers and his following statements are worthy of repetition.

with the same material providing the main construction stone of the round-arched doorway, suggesting that this too was largely rebuilt (Figure 7.6). It proved impossible to read the lineation of any of the granite blocks. The arch to the doorway is hewn from a single block of granite with certainly one of the two imposts of phyllite. The masonry in the walls of the tower, at least in its lower visible levels, tends to be created only with BH blocks of phyllite/schist, and only in very rare instances are edge bedded blocks inserted in ‘Romanesque’ style to assist bonding (see section 2.13.3), in circumstances which may indicate repairs. Granite boulders are also present in the walls.

‘… the most enigmatic of medieval buildings anywhere …’ ‘There is general agreement that the main period of their construction was between the start of the tenth century and the end of the twelfth’ ‘There is also agreement that they were, first and foremost, the bell-houses of church sites … symbols of wealth of monasteries …occasionally treasuries in which relics and other valuables were stored …’ O’Keeffe (2004, 11)

This description of the Glendalough tower is by no means intended to be complete. Rather it introduces some of the features of stonework which might be utilized to determine the period of a tower’s first construction. These are: i) The doorway – doorways vary in form and as indicated by O’Keeffe (2004, 68) it is probably possible to date these by comparison with equivalent forms in Irish churches. Thus, if a doorway possessed an architraved surround with cut backs, well illustrated in the Scottish example of Brechin (see section 3.3.5), they would be of comparable late ‘Patterned’ age. A ‘Patterned’ construction date should be linked to any doorway with BVFIA/BH jambs. In contrast, O’Keeffe described the details of a number of post mid-11th century doorways, all of which should possess BH jambs. ii) The windows – windows may again be dated according to the style of their jambs as revealed at

Only the masonry of Glendalough Round Tower (T 123 968) (Figure 7.2) was examined by the author in sufficient detail for a construction date to be determined. Captured in unusual strong oblique lighting and scrutinized with powerful binoculars, seven of the eight windows examined throughout the tower’s height exhibited jamb stones set in the ‘Patterned’ style (Figures 7.3 to 7.5). For these seven windows, largely created with blocks of phyllite/schist, all visible stones were orientated BVFIA/BH in the style of the period (in several instances exhibiting long and short characteristics). Only the southwest lowest window was fashioned entirely with later BH orientated stones. In a few instances, later replacement stones of granite were incorporated in certain windows, 177

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND

Figure 7.3 The west window at the top of the Glendalough round tower is constructed of blocks of phyllite which are set in ‘Patterned’ style. The jambs, each of three stones, are placed in long and short setting and orientated BVFIA, BH, BVFIA.

Figure 7.4 The south window at the top of the Glendalough tower is similar to the west window (Figure 7.3), but the right jamb appears to have three BH stones between the BVFIA stones.

Figure 7.5 A further example of the windows in Glendalough tower, this the east window illuminating the level below the top floor, with single BVFIA jamb stones. The lintel stone attracting the lichen may be of granite.

Figure 7.6 The doorway to the Glendalough round tower is largely reconstructed of granite (which attracts lichen). Any lineation in these stones is impossible to determine. 178

CHAPTER SEVEN Glendalough. O’Keeffe (2004, 77) gave details of the various common window shapes. iii) The masonry – and the possibility of distinguishing ‘Romanesque’ from ‘Patterned’ workmanship on the basis of the use of edge bedded stones inserted to assist bonding.

(dated c. 3500 BC) in Papa Westray, Orkney (HY 483 517). The ‘Patterned’ period cannot, however, have been an interlude of extreme and universal virtue, when valuables could remain unprotected, and reference has been made above to church attacks both from Vikings and others. Some method of protection would surely have been required.

In the instance of Glendalough tower, the windows and masonry would seem to indicate a ‘Patterned’ date of original construction. Of the limited number of round towers in Ireland viewed (but not examined in any detail) by the present author, the majority appear to be of ‘Romanesque’ character. As indicated by O’Keeffe (2004, 88), more than half of the towers in Ireland are probably undatable because they are too fragmentary; to this comment may be added, and what remains may have suffered extensive alteration.

The unusual shape of the typical west doorway to the early ‘Patterned’ church or round tower, although often described, has achieved little analytical comment. Its upward taper towards the heavy lintel generally receives note, but the variation in width between the jambs, from exterior to interior, is not acknowledged. Measurements of this change in width (recorded to the nearest 5mm.) were made for a small number of churches and are presented here in Table 7.1. Unexpectedly, for the 16 doorways examined there is little clear pattern, and virtually the only door that could not be fitted was one with parallel sides. No doubt settlement over a long period has influenced the measurements but it would appear that every door was constructed to purposefully fit the enclosing aperture.

7.1.7 The west doorway – ‘A priest in residence?’ It might be argued that the western entrances to the early ‘Patterned’ churches, as those observed in Ireland and elsewhere, were open and unprotected – affording a welcome to all wishing to follow the faith. There is no evidence of any form of door or door fastening, and yet, typical door fixings and attachments are known to have been in use well before this period of church building. The Romans made extensive use of doors comparable to those of today, and in the British Isles, for example, door fastenings can be observed at the Knap of Howar site

A greater degree of security would have probably been gained where the aperture was wider on the exterior. For this form of aperture, the door, presumably a composite, heavy wooden structure would have been pulled into position. If tightly fitting, in the absence of any external ornamentation, any assailant would have had the difficult

Table 7.1 West doorway widths for various early Irish churches. Measurements are recorded to the nearest five millimetres Church Site

Grid Reference

Aperture of West Doorway At Top

Ardagh,Longford1 Clonamery, Kilkenny2 Fore, Westmeath Friar’s Island, Killaloe, Clare3 Glendalough, St Mary, Wicklow Inishmore, St Benan, Galway Inishmore, T. Macduagh, Galway Inisheer, Kilgobnet, Galway Kill of the Grange, Dublin Killoughternane, Carlow Kilree, Kilkenny Labbamolag(g)a (small) Co. Cork Oughtmama (west), Clare Ratass, Kerry4 St John’s Point, Co. Down5 Tuamgraney, Clare

N 204 686 S 658 354 N 510 704 R 701 728 T 122 968 L 884 071 L 823 104 L 975 027 O 226 273 S 777 542 S 497 410 R 764 176 M 304 078 Q 853 141 J 528 338 R 637 830

Wider By mm. Interior 15 Interior 10 Interior 50 Interior 5 Interior 10 Exterior 35 Exterior 10 Exterior 20 Exterior 15 -Neither 0 --Exterior 10 Exterior 5 Exterior 20

Notes: 1 Interior extensively repaired. 2 Interior of west gable wall thickened at later date to take bell tower. 3 Rebuilt. 4 Interior rebuilt. 5 Extensively repaired. 179

At Bottom Wider By mm. -Exterior 10 Interior 45 Neither 0 Exterior 10 Exterior 15 Exterior 10 Exterior 10 -Exterior 40 Exterior 20 Exterior 30 Exterior 20 Interior 10 Neither 0 Exterior 30

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND task of levering the door out of its position by means of the extreme edges. A door pushed into a doorway opening from a wider interior would necessitate internal support or wedges to maintain security. In either situation the building inside would require the presence of a person or persons to try to preserve this protection. The vertical taper to the doorways would enable doors to be presented correctly to the aperture, more easily and rapidly.

raised in the text. Crow-stepped (corbie) gables, for instance, were briefly examined in northernmost Scotland to try to determine precisely when their introduction might first have occurred (see section 3.3.6). Hoey and Thurlby (2004) made a survey of ‘Romanesque’ vaulting (although it is clear that they were unable to visit each cited locality). The work is confined to the period 1066 to around 1170, so that it fails unfortunately to cover any possible work of the ‘Patterned’ period of prior to 1066. A number of churches possess vaulted crypts which were constructed during the ‘Patterned’ period (such as Repton, and probably Wing) and further work would be useful to determine the considered date of the first introduction of such crypts and the existence of possible changes in their style or application.

Without further study this explanation as to the variation in doorway width must remain one largely of conjecture. Past views have proposed that, at times of impending danger, valuables from the church were transferred to the adjoining round tower and, therefore, a church door was unnecessary. The round towers typically possess a similar tapered doorway and it would appear that both forms of building would have had the same level of entrance security – the towers’ protection being enhanced by the doorway elevation from the ground.

Although other possible aspects of architectural detail afford similar opportunities for study, the variation in window development over time is a topic in serious need of examination. Shapes, splay variables and amounts, positions, structures, chamfering and other details, all change considerably. Just the placing of these changes into some recognised order, for example, would particularly assist in helping to determine the dates of many of Scotland and Ireland’s ruined churches.

7.1.8 Earlier and later stonework styles Although some elements of Norman or ‘Romanesque’ stone emplacement practices have been referred to (see section 2.13), little comment has been made with regard to Roman practices apart from the statement that Roman buildings appear to be ‘reliant on BH orientations for their stonework’. The present author has probably examined more stonework in Roman standing and ruined buildings throughout the one-time empire than most, but a far more comprehensive study is required. ‘Romanesque’ workmanship also requires far more extensive study. Still more recent, medieval and later styles also require examination. However, as one approaches more modern practices, the work appears to become increasingly casual with regard to stone bedding orientation.

7.2 Conclusions 7.2.1 Introduction An analysis of the findings of this study provides certain insights into the employment of vertically bedded stonework as it occurs in early ecclesiastical buildings of Britain and Ireland. It is now evident that craftsmen working in stone in the period prior to that of the ‘Romanesque’ architectural styles used hewn stones in unusual and distinctive ways. In England, these stone working traditions are observed particularly in structural features like quoins, pilaster-strips and the jambs of arches (Potter 2005b). The styles adopted involved orientating the bedding, stratification, or any lineation, in many of the stones in these structures, vertically. This was probably mainly undertaken for decorative reasons, and where rock types permitted, the wall masonry has also been shown to display forms of decoration.

The occasional use of vertical bedding orientation of stone of contrasting colour to the main church fabric (‘rustication’) as occurs in Scotland (see section 3.3.2) and in Ireland (see section 4.4.3) during the period from the mid-17th century has not been observed in England. However, in England, vertical orientation may very occasionally be observed on the edge stones of buttresses. An example of such an occurrence may be seen in the east buttress to the north aisle of Sullington Church, West Sussex (TQ 098 131). Buttresses are extraordinarily difficult to date with certainty but this one may have been added as recently as in the 1870s. The application of vertically bedded stones is to enable the edge or quoin of the buttress to be completed with the smooth face of a vertical stone rather than the broken edge of a thicker BH stone.

In Scotland and Ireland, where the rocks were less compliant, pilaster-strips were not used to enhance the decoration of the period. In these countries too, the term ‘Patterned’ is now offered, instead of the English, ‘Anglo-Saxon’ to describe these styles of workmanship; although the evidence suggests that any time differences between the regions is relatively slight. The more extensive study of early Irish churches (Chapter 4) had one principal objective – to illustrate the importance of the presence of particular stone working styles in both ‘Patterned’ and the subsequent ‘Romanesque’ architecture in that country. It was not intended to give a detailed analysis of the structure of any particular church. Such an analysis would have required a more detailed examination of the stonework involving the

7.1.9 Architectural detail In the course of the present work it has become apparent that particularly in Scotland and Ireland, where possibly early church architecture has been less extensively studied than in England, aspects of this work require further study. In some instances these aspects have been 180

CHAPTER SEVEN

Figure 7.7 Lemanaghan Oratory, Offaly (N 176 270), when visited in 2005 was being rebuilt. Note the pile of stones beside the north wall of the church.

Figure 7.8 Pyrford church, Surrey (TQ 040 583) is principally constructed of local ferruginously-cemented gravel, but modern replacement stones in structural situations such as this window consist of starkly contrasting white Upper Jurassic, Portland Stone.

181

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND careful cleaning of certain stones. No attempt was made to remove lichen or other masking materials. Generally two, with a maximum of three, visits to each site, totalling no more than six hours of scrutiny, provided sufficient information to determine specific stonework styles. It is the author’s intention to provide amplified descriptions of some of these structures (and churches) elsewhere. This work recognises only their presence; it is intended that it will encourage others to seek similar features in other Irish church sites. Only the most clearly discernable examples of vertically orientated bedding in quoins, antae and arch jambs have been included. Various less obvious examples exist or remain doubtful, in many instances these require extensive stone cleaning and quality oblique natural light to assist in their ultimate interpretation. Finally, the Irish studies tried (Chapter 5), without any real success, to interpret the masonry studies of Ó’Carragáin (2005a; 2005b): they also provided a new interpretation for the presence of antae in the Irish churches (Chapter 6).

A strong case clearly exists for standard procedures across the whole of the British Isles. The subjects of conservation and rebuilding of ecclesiastical sites should be debated in full prior to any decision as to which procedures should be followed. 7.2.3 Great Britain, Ireland and beyond The recognised strength of Celtic Christianity in Ireland provided the present author with the necessary incentive to investigate Irish monastic and religious sites for evidence of ‘Patterned’ style workmanship. It seemed possible that Ireland might provide the early source for the distinctive styles of church construction that embrace this style. Although clearly evident in Ireland, the country of origin within the British Isles remains uncertain. Irish archaeologists remain reluctant (or the evidence is insufficient) to suggest that stone built Christian sites of worship existed in Ireland as early as they appear to have existed in England (see section 4.1). In Scotland too, stone ecclesiastical sites have only been identified from the later phases of the ‘Patterned’ period.

7.2.2 Conservation and preservation of early ecclesiastical buildings

To look for earlier examples of ‘Patterned’ style workmanship than the earliest stone church sites in England it may be necessary to look to mainland Europe, and even to the home of Christianity in the Middle-East.

Although those concerned with the conservation of ecclesiastical buildings would defensively argue that random stone replacement in walls never occurs; the present author has encountered various instances where piles of stone represent the remains of a church wall (Figure 7.7) which is eventually to be reconstructed. In some instances, the wall may have collapsed, but in others, the fabric has been taken down piecemeal prior to rebuilding. It might correctly be suggested that this procedure is no different than the similar actions of predecessors over the 900 years or so of an early building’s previous history. Indeed, where early stone types are lithologically distinctive, the re-use of stone from such a pile has permitted the present author to identify an earlier, as Anglo-Saxon, origin to certain churches in England (Potter, 2001a). In Scotland and Ireland, in particular, in more recent times, a thin layer of black plastic or mastic is placed between old (?original) and new (rebuilt) wall; but on occasions this appears to be absent. In these countries such distinction is possibly more critical, for rock lithologies tend to show greater similarities – hence works such as that of Ó’Carragáin (2005a), where reliance moved instead to masonry styles. In England, whether by design or accident, modern replacement stones are currently often of an unfortunate, vividly contrasting, rock lithology (Figure 7.8). This work has shown that, if a wall is to be rebuilt, not only do rocks which are to be re-used have to be identified by their exact position in a wall; also their exact orientation is required: this, of course, if a precise replica is required of the wall. If the early photographs of the Friar’s Island church (as it once stood on the island) are viewed it is clear that the church (see section 4.5.8) was then far more dilapidated than it is today. The church as it stands at Killaloe, is mixture of copied old work and new, with no obvious distinction; making interpretation unreliable. 182

GLOSSARY A number of words in this work are not used with great regularity in archaeological literature. They are in some instances defined when they first appear in the document. These may be defined again here to enable the casual reader to more readily appreciate their meaning. Geological terms used in the work which relate to units of rock are placed according to their age in the stratigraphical column provided at the end of this Glossary.

GREYWACKE: an unsorted, rapidly deposited, sedimentary rock of mixed mineralogical and textural composition. HAEMATITE: also spelt hematite, iron oxide, Fe2 O3. IGNEOUS: a rock solidified from a molten, or partially molten, state. JOGGLED: stones fitted into a wall and partially interlocked in such a way (joggled joints) as to prevent them sliding at least in one direction. KEEILL: an Isle of Man term for a small chapel or hermitage. LITHOLOGY: the appearance, grain size and composition of a particular rock; generally with reference only to sedimentary rocks. LONDON BASIN: the downfolded (that is, synclinal) structure created between the Chalk of the Chilterns and the North Downs. Rocks inside and lining the Basin are all more recent in age geologically than the Chalk. MASSIVE bedding: where the beds in a sedimentary rock remain homogeneous over a considerable thickness (typically several metres) and stratification or bedding is apparently absent. As opposed to thinly bedded. METAMORPHIC: a rock formed under the influence of a considerable amount of heat and/or pressure, typically with some resultant mineral recrystallisation. MICA SCHIST: a relatively medium grained metamorphic rock in which the intense heat and pressures involved have developed the sheet structured mineral mica; the abundance of this mineral imposes a foliation (or schistosity) on the rock. OOLITIC: a rock composed of millions of ooliths (literally, ‘egg’ ‘stone’); such as an oolitic limestone, where the ooliths are largely composed of calcium carbonate. Ooliths in a limestone are created by the deposition of Ca CO3 on wave disturbed, and therefore moving, minute particles in a shallow sea. PAN: (in particular iron pan) a layer of hydrated iron oxides/iron oxides which may develop approximately at the level of the surface of the water table in permeable rocks like gravel or sand due to a degree of evaporation PERPENS: a masons’ term for the vertical mortar or cement joints between bricks or stones in a wall. The horizontal joints are referred to as the beds. PHYLLITE: a fine grained, low grade metamorphic rock with a well developed fine schistosity (which tends to give the flat surfaces of the rock a smooth silky sheen). QUARTZITE (quartzitic): a sandstone with a silica cement. A sarsen is a variety of quartzite. SARSEN: a silicified sand or silcrete, probably formed at the surface or slightly below the surface, of sand deposits, under certain climatic conditions. Following extensive weathering these may be found as large irregular boulders (sarsens) in superficial deposits. SCHIST: a relatively medium grained metamorphic rock in which the intense heat and pressure develops a mineral foliation or schistosity. See mica schist. SCORIACEOUS: a term applied to frothy, gas rich, basaltic lavas, which tend when cold to display numerous small holes or vesicles.

AMMONITE: the ammonites were a, now extinct, fossil group of molluscs, typified by a planar coiled chambered shell. They were abundant especially in the Jurassic seas. ARRIS: the sharp, generally straight, edge between two planar surfaces; and a term used particularly when working stone. ASHLAR: squared hewn stone or stones, or masonry constructed of such stones. The shape of such blocks is often determined by its jointing or incipient jointing. It should be noted that boulders used in megalithic quoins, for example, need not necessarily be ashlar in form. BASALT (basaltic): a volcanic (or extrusive) igneous rock or lava, which is fine grained and dark in colour. BEDROCK: the solid rock underlying the much more recently formed, and typically unconsolidated, superficial deposits. CHERT: microcrystalline silica; in sedimentary rocks it may be chemically precipitated to form thin seams or nodules. Flint is a distinctive variety of chert which occurs in the Chalk. CLEAVAGE: in rocks. Cleavage in a rock creates a tendency for the rock to break along thin, parallel, and closely spaced, layers. The layers, known as the cleavage, are generally imparted as a result of pressure during metamorphism and normally involve some degree of mineral recrystallisation. CORBIE: A term used in Scotland for crow-stepped (see below). CROW-STEPPED: squared stones set like steps up the edge of a gable or gateway. ESKER: a long ribbon like hill composed of fluvioglacial sands and gravels originally deposited by rivers flowing beneath an ice-sheet or from the front of a retreating ice-sheet. The Clonmacnoise area in Ireland is famous for their occurrence. FISSILE (fissility): easily broken or split into thin layers. In a rock this is normally due to the rock possessing a marked cleavage. FLUVIOGLACIAL: created by melt waters from glaciers. GLAUCONITE (glauconitic): a greenish complex silicate mineral that is often found in sedimentary rocks that have been deposited in shallow marine conditions. The mineral tends to impart a green colour on the sediment. GNEISS: a coarse grained, generally banded, metamorphic rock formed as a result of very extensive heat and pressure.

183

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND SEPTARIA (septarian nodules): in certain clay deposits (such as the London Clay), calcium carbonate (sometimes iron rich) concretions may have been formed, normally of cobble size, but occasionally up to 2m. across. In these concretions irregular polygonal dehydration cracks (from which the name septaria) may be present. SILCRETE: surface or very near to surface deposit rich in silica particularly formed in warm to subtropical climates. See also sarsen. SKEUOMORPH (skeuomorphic): an object which in its shape and decoration is thought to resemble or copy a form or technique made in or of another different material. SPECULAR: of flat, tabular flakes, resembling mica. SUPERFICIAL: typically unconsolidated, surface deposits of recent geological age, such as river gravel.

THINLY BEDDED: As opposed to massively bedded. Sedimentary rocks which exhibit frequent minor changes in lithology, typically displayed as layers of bedding. TRAVERTINE: See Tufa. TUFA: a rock of cavernous, precipitated calcium carbonate. Tufa is relatively light in weight, and normally soft when first precipitated so that it can be readily be cut or sawn into blocks. In England it was worked extensively by the Romans as a building stone. The rock hardens on exposure, when it is better described as travertine. VESICLES (vesicular): small holes created and preserved in lava by the presence of excessive gases during cooling. See also scoriaceous.

184

A chrono-stratigraphic column for the rocks of Britain cited in this work. Details of the rocks referred to in Ireland are given in Table 4.1 and the text CHRONO-STRATIGRAPHIC UNITS ________________________________

APPROX. AGE (Millions of Years)

ROCK UNIT EXPLOITED

ERA SYSTEM SERIES -------------------------------------------------------------------------------------------------------------------------------------------Cainozoic Quaternary Holocene A. Ferruginously-cemented sands and gravels and B. Bunter Quartzites Pleistocene (2) C. Travertine D. Sarsen, Hertfordshire Puddingstone, Leziate Stone – all being silcretes ------------------------------------------------------------------------------------------------------------------------Palaeogene Eocene 40 A. Quarr Stone (from the Isle of Wight) B. Septarian nodules from the London Clay ------------------------------------------------------------------------------------------------------------------------------------------Mesozoic Cretaceous Upper 85 A. Lavant Stone ) Varieties of B. Totternhoe Stone ) hard Chalk Lower

110

A. Upper Greensand B. Folkestone Formation Sands ) Lower C. Sandgate Formation: Bargate Member) D. Hythe Beds ) Greensand E. Car(r)stone from Norfolk area F. Ashdown Beds Sandstone 145 G. Viviparus Limestone (‘Purbeck Marble’) -------------------------------------------------------------------------------------------------------------------------Jurassic Upper 155 A. Oxford Clay Formation (Ancholme Group) Middle

170

A. Barnack Stone (a variety of shelly oolitic limestone) --------------------------------------------------------------------------------------------------------------------------------------------Upper Permian 260 A. Stornaway Beds (sandstones) Palaeozoic -------------------------------------------------------------------------------------------------------------------------Carboniferous Mississipian 345 A. Carboniferous Limestone -------------------------------------------------------------------------------------------------------------------------Devonian Middle 390 A. Middle Old Red Sandstone --------------------------------------------------------------------------------------------------------------------------------------------Lower Silurian 425 Palaeozoic -------------------------------------------------------------------------------------------------------------------------Ordovician 460 -------------------------------------------------------------------------------------------------------------------------Cambrian 520 A. Manx Slate Group B. Swithland Slate --------------------------------------------------------------------------------------------------------------------------------------------Precambrian 560 A. Dalradian (Supergroup) --------------------------------------------------------------------------------------------------------------------------------------------ROCKS FROM FRANCE: Calcaire Grossiere – an Eocene limestone from the Paris Basin Marquise Oolite – a Middle Jurassic Oolite from the Boulogne area Caen Stone – a Middle Jurassic limestone from Normandy

185

REFERENCES Note: Archaeological Inventories of various Irish Counties. Over the period 1992 to the Present, a series of inventories of the various archaeological sites in Irish Counties have been produced. Currently about half of the Irish Counties have been covered. The works are published by the Stationery Office, Dublin (Duchas, The Heritage Society). Aldsworth, F.G. and Harris, R. 1988. The tower and ‘Rhenish helm’ spire of St. Mary’s Church, Sompting. Sussex Archaeological Collections, 126, 105-144. Allen, J.R. and Anderson, J. 1903. The Early Christian monuments of Scotland. 2 vols. Society of Antiquaries of Scotland, Edinburgh. Allen, J.R.L. 2004. Carrstone in Norfolk buildings. Distribution, use, associates and influences. British Archaeological Reports, British Series No. 371, BAR Publishing, Oxford. Allen, J.R.L., Pearson, A.F. and Fulford, M.G. 2001. Branodunum on the North Norfolk coast: a local origin for the building material. Britannia, 32, 271275. Allen, J.R.L., Rose, E.J. and Fulford, M.G. 2003. Re-use of Roman stone in the Reedham area of Norfolk: intimations of a possible ‘lost’ Roman fort. Britannia, 34, 129-141. Anderson, F.W. and Quirk, R.N. 1964. Note on the Quarr Stone, 115-117. In: E.M. Jope, The Saxon stone industries in southern and midland England. Medieval Archaeology, 8, 91-118. Anon. 1891. (Descriptions of excursions in County Kerry and elsewhere). Journal of the Royal Society of Antiquaries of Ireland, 21, 609-631. Bagshaw, S. 1998. Early medieval church to tractor shed, building survey at Leonard Stanley chapel. Glevensis, 31, 3-12. Barrow, G.L. 1979. The round towers of Ireland. A study and gazette. Academy Press, Dublin. Barry, T.B. 1987. The Archaeology of Medieval Ireland. Methuen, London. Berger, R. 1992. 14C dating mortar in Ireland. Radiocarbon, 34, 880-889. Berger, R. 1995. Radiocarbon dating of early medieval Irish monuments. Proceedings of the Royal Irish Academy, 95C, 159-174. Biggar, F.J. 1896. Cruach MacDara, off the coast of Connamara: with a notice of its church, crosses, and antiquities. Journal of the Royal Society of Antiquaries of Ireland, 26, 101-112. Bolton, J.L. 1980. The medieval English economy 11501500. Dent, London. Bone, D.A. and Bone, A.E. 2000. Lavant Stone: a late Roman and medieval building stone from the Chalk (Upper Cretaceous) of West Sussex. Proceedings of the Geologists’ Association, 111, 193-203. Bradley, J. 1980-81. St Patrick’s Church, Duleek. Riocht na Midhe, 7 (1), 40-51. Brash, R.R. 1857-60. The round tower of Abernethy, with drawings. Proceedings of the Society of Antiquaries of Scotland, 3, 303-319. Brash, R.R. 1875. Ecclesiastical Architecture of Ireland. W.B.Kelly, Dublin.

Bridgland, D.R. 1994. Quaternary of the Thames. Geological Conservation Review Series, No. 7, Chapman and Hall, London. Brown, G.B. 1903. The Arts in early England: Ecclesiastical Architecture in England from the conversion of the Saxons to the Norman Conquest. John Murray, London. Brown, G.B. 1925. The Arts in early England: AngloSaxon Architecture. John Murray, London. Browne, E.A. 1907. Great buildings and how to enjoy them: Norman architecture. A. and C. Black, London. Byrne, M.J. 1910. Rattoo. Journal of the Cork Historical and Archaeological Society, 16, 182-193. Byrne, M.J. 1911. Rattoo. Journal of the Cork Historical and Archaeological Society, 17, 17-25. Cameron, N. 1994. St Rule’s Church, St Andrews, and early stone-built churches in Scotland. Proceedings of the Society of Antiquaries of Scotland, 124, 367-378. Champneys, A.C. 1910. Irish Ecclesiastical Architecture: with some notice of similar or related work in England, Scotland and elsewhere. G. Bell & Sons, London. Chancellor, F. 1904-5. Inworth Church. Transactions of the Essex Archaeological Society, n.s. 9, 357-360. Clapham, A.W. 1930. English Romanesque architecture before the Conquest. Clarendon Press, Oxford. Clapham, A.W. 1934. English Romanesque architecture after the Conquest. Clarendon Press, Oxford. Cleary, R.M. 2000. Labbamologa, Co. Cork. Journal of the Cork Historical and Archaeological Society, 105, 27-48. Clifton-Taylor, A. 1972. The pattern of English building. Faber and Faber, London. Cochrane, R. 1911-12. Historical and descriptive notes of the ecclesiastical remains at Glendalough, Co. Wicklow. Commissioners for Public Works in Ireland, Dublin. Cochrane, R. 1912. Notes on the structures in the County of Cork vested in the Board of Works for preservation as ancient monuments. Journal of the Cork Historical and Archaeological Society, 18, 1-25. Cooke, T.L. 1842-3. Autumnal rambles about New Quay, County Clare. Galway Vindicator. Crawford, B.E. (ed.) 1988. St Magnus Cathedral and Orkney’s Twelfth-century Renaissance. Aberdeen. Crawford, H.S. 1908. Ardpatrick, County Limerick (note). Journal of the Royal Society of Antiquaries of Ireland, 38, 75-76. Cruden, S. 1986. Scottish medieval architecture. J. Donald, Edinburgh. Cuffe, P. 1963. The Priory of Duleek. Riocht na Midhe, 3 (1), 29-32.

186

REFERENCES Cuffe, P. 1965. History of Duleek: section II – ecclesiastical; AD.432 – AD.1172. Riocht na Midhe, 3 (3), 187-200. Cunningham, G. 1976. Roscrea and District: monuments and antiquities: illustrated guide. Parkmore Press, Roscrea. de Paor, L. 1967. Cormac’s chapel: the beginnings of Irish Romanesque, 133-145. In: E. Rynne (ed.), North Munster Studies, Limerick. Dietrichson, L. and Meyer, J. 1906. Monvmenta Orcadia: the Norsemen in the Orkneys and the monuments they have left. Williams and Norgate, London. Dixon, R. 2005. Field meeting report – the geology of some East Anglian Crag churches: 7th May 2005; leader Prof. John F. Potter. Magazine of the Geologists’ Asssociation, 4 (3), 8-9. Duignan, M. 1951. Early monastic site, Kiltiernan East Townland, Co. Galway. Journal of the Royal Society of Antiquaries of Ireland, 81, 73-75. Dunraven, (Lord E.R.W.). 1875. Notes on Irish Architecture. Vol. 1. Bell, London. Edwards, N. 1996. The Archaeology of Early Medieval Ireland. Routledge, London. Fahey, J. 1901. Some antiquities in the neighbourhood of Oranmore and Kilcolgan, County Galway. Journal of the Royal Society of Antiquarians of Ireland, 31, 228235. Fahey, J. 1904. Kilmacduagh and its ecclesiastical monuments. Journal of the Royal Society of Antiquaries of Ireland, 34, 220-233. Fanning, T. 1977. An ogham stone and cross-slab from Ratass Church, Tralee. Journal of the Kerry Archaeological and Historical Society, 10, 14-18. Fanning, T. 1981. Excavation of an Early Christian cemetery and settlement at Reask, Co. Kerry. Proceedings of the Royal Irish Academy, 81C, 67172. Fawcett, R. 2002. Scottish medieval architecture: architecture and furnishings. Tempus, Stroud. Fernie, E.C. 1983. The architecture of the Anglo-Saxons. Batsford, London. Fernie, E.C. 1986. Early church architecture in Scotland. Proceedings of the Society of Antiquaries of Scotland, 116, 393-411. Fernie, E.C. 1988. The church of St Magnus, Egilsay, 140-161. In: B.E. Crawford (ed.) 1988, St Magnus Cathedral and Orkney’s Twelfth-century Renaissance. Aberdeen. Fernie, E.C. 2000. The architecture of Norman England. Oxford University Press, Oxford. Fitzpatrick, E. and O’Brien, C. 1998. The medieval churches of County Offaly. Government of Ireland, Dublin. Fletcher, E.G.M. and Jackson, E.D.C. 1945. ‘Long and short’ quoins and pilaster strips in Saxon churches. Journal of the British Archaeological Association, 3rd Ser., 9, 12-29. Freke, D.1995. The Peel Castle dig. The Manx Museum, Douglas, Isle of Man. Garton, T. 1981. A Romanesque doorway at Killaloe. Journal of the British Archaeological Association, 134, 31-57.

Geological Survey of Ireland, 2006. Bedrock geological map of Ireland, 1:500,000. Gifford, J. 1992. The buildings of Scotland: 5. Highland and Islands. Penguin, London. Gilbert, E. 1946. New views on Warden, Bywell and Heddon-on-the-wall churches. Archaeologia Aeliana, 4th Ser., 24, 157-162. Gleeson, D.F. 1951. Churches in the deanery of Ormond. North Munster Antiquarian Journal, 6 (3), 96-107. Gwynn, A. and Hadcock, R.N. 1970. Medieval Religious Houses: Ireland. Longman, London. Hamilton, T. 1903-4. The repair of Dromacoo old church. Journal of the Galway Archaeological and Historical Society, 3, 116-125. Hamlin, A. 1984. The study of early Irish churches, 117126. In: P. Ní Chatháin and M. Richter (eds), Irland und Europa, Die Kirche im Frühmittelalter. KlettCotta, Stuttgart. Hancock, P.L. and Skinner, B.J. (eds) 2000. The Oxford companion to the Earth. Oxford University Press, Oxford. Harbison, P. 1970. How old is Gallarus Oratory? A reappraisal of its role in early Irish architecture. Medieval Archaeology, 14, 34-59. Harbison, P. 1972a. Some Romanesque heads from County Clare. North Munster Antiquarian Journal, 15, 3-7. Harbison, P. 1972b. The building-history of Ratass Church, Tralee. Journal of the Kerry Archaeological and Historical Society, 5, 5-13. Harbison, P. 1982. Early Irish churches, 618-629. In: H. Löwe (ed.), Die Iren und Europa im früheren Mittelalter I1. Klett-Cotta,Stuttgart. Harbison, P. 1992. Guide to national and historical monuments of Ireland. Gill and Macmillan, Dublin. Harbison, P. 1995. Architectural sculpture from the twelfth century at Ardmore. Irish Arts Review, xi, 96102. Harbison, P. 1999. The Clones sarcophagus – a unique Romanesque style monument. Archaeology Ireland, 13 (3), 12-16. Hare, M. and Hamlin, A. 1986. The study of early church architecture in Ireland: an Anglo-Saxon viewpoint, 131-145. In: L.A.S. Butler and R.K. Morris (eds), The Anglo-Saxon church: Papers on history, architecture and archaeology in honour of Dr H M Taylor. CBA Research Report 60. Hart, S. 2003. The round church towers of England. Lucas Books, Ipswich. Henry, F. 1970. Irish art in the Romanesque period (1020-1170 AD). London. Herity, M. 1984. The layout of Irish early Christian monasteries, 105-116. In: P. Ni Chatháin and M. Richter (eds), Ireland and Europe: the early church. Klett-Cotta, Stuggart. Hinton, D.A. 2002. Debate: the dating of ferruginouslycemented gravel as building material. Landscape History, 24, 121-122. Hoey, L.R. and Thurlby, M. 2004. A survey of Romanesque vaulting in Great Britain and Ireland. The Antiquaries Journal, 84, 117-184. 187

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND Holland, C.H. (ed.) 1981. A geology of Ireland. Scottish Academic Press, Edinburgh. Hurley, V. 1982. The Early Church in the South-West of Ireland: Settlement and Organisation, 297-332. In: S.M. Pearce (ed.), The Early Church in Western Britain and Ireland: Studies presented to C. A. Ralegh Radford, British Archaeological Reports, British Series 102. Hussey, R. 1858. St. Mildred’s, Canterbury. Archaeologia Cantiana, 1, 143-149. Jackson, E.D.C. and Fletcher, E.G.M. 1949. Further notes on ‘long’ and ‘short’ quoins in Saxon churches. Journal of the British Archaeological Association, 3rd Ser., 12, 1-18. Jervise, A. 1857-60. Remarks on the round tower of Brechin. Proceedings of the Society of Antiquaries of Scotland, 3, 28-35. Jones, C. 2005. The Burren and the Aran Islands: exploring the archaeology. Collins, London. Jope, E.M. 1974. The Saxon building stone industries in south and midland England. Medieval Archaeology, 8, 91-118. Laing, L. 1977. The archaeology of Late Celtic Britain and Ireland c. 400 – 1200 AD. Book Club Associates, London. Lalor, B. 1999. The Irish Round Tower: origins and architecture explored. Collins, London. Laming, D. 2003. Churches built of gravel?? A field meeting to some churches in the Reading/Slough area, 10th May 2003: leader John Potter. Magazine of the Geologists’ Association, 2 (4) 9-10. Leask, H.G. 1929. Further notes on the church [on Friar’s Island, Killaloe]. Journal of the Royal Society of Antiquaries of Ireland, 59, 25-28. Leask, H.G. 1930. The church of St Lua, or Molua, Friar’s Island, Co. Tipperary, near, Killaloe. (Further notes). Journal of the Royal Society of Antiquaries, 60, 130-136. Leask, H.G. 1955. Irish Churches and Monastic Buildings I. The first phases and the Romanesque. Dundalgan Press, Dundalk Leask, H.G. 1958. Irish Churches and Monastic Buildings 11. Dundalgan Press, Dundalk. Leask, G.M. 1977. Glendalough, Co. Wicklow: National monuments vested in the Commissioners of Public Works. Dublin. Livett, G.M. 1893. Early-Norman churches in and near the Medway Valley. Archaeologia Cantiana, 20, 137154. Lynch, P.J. 1909. Cromleacs in Co. Limerick. Journal of the North Munster Archaeological Society, 1 (1), 3036. (It is possible that the end note to the article should be attributed to Anon). Macalister, R.A.S. 1909. The memorial slabs of Clonmacnois, King’s County. The Royal Society of Antiquaries of Ireland, Dublin. Macalister, R.A.S. 1929. On some excavations recently conducted on Friar’s Island, Killaloe. Journal of the Royal Society of Antiquaries of Ireland, 59, 16-24. MacDonald, A. 2003. The ‘Cathedral’, Temple Kelly and Temple Ciarán: notes from the Annals, 125-135. In:

H.A. King (ed.), Clonmacnoise Studies, Vol. 2, Seminar Papers, 1998, Dúchas, Dublin. McNab, S.L. 1987. The Romanesque sculptures of Ardmore Cathedral, Co. Waterford. Journal of the Royal Society of Antiquaries of Ireland, 117, 50-68. McNeill, C. and Leask, H.G. 1920. Monaincha, Co. Tipperary, historical notes: architectural notes. Journal of the Royal Society of Antiquaries of Ireland, 50, 19-35. McNeill, J.T. 1974. The Celtic churches: a history A.D. 200 to 1200. University of Chicago Press, Chicago. McRoberts, D. 1968. Material destruction caused by the Scottish Reformation. Innes Review, 19, 126-172. Manning, C. 1995a. Early Irish monasteries. Country House, Dublin. Manning, C. 1995b. Clonmacnoise Cathedral – the oldest church in Ireland? Archaeology Ireland, 9 (4), 30-33. Manning, C. 1996. The nave of Glendalough Cathedral. IAPA Newsletter, Bulletin of the Irish Association of Professional Archaeologists, 9 (4), 30-33. Manning, C. 1998. Clonmacnoise Cathedral, 56-86. In: H.A.King (ed.), Clonmacnoise Studies, Vol. 1, Seminar Papers 1994, Dúchas, Dublin. Manning, C. 2000. References to church buildings in the Annals, 37-52. In: A.P.Smythe (ed.), Seanchas: studies in early and medieval Irish archaeology, history, and literature in honour of Francis J. Byrne. Four Court Press, Dublin. Manning, C. 2002. A puzzle in stone; the cathedral at Glendalough. Archaeology Ireland, 16 (2), 18-21. Manning, C. 2003. Some early masonry churches and the round tower at Clonmacnoise, 63-95. In: H.A. King (ed.), Clonmacnoise Studies, Vol. 2, Seminar Papers 1998, Dúchas, Dublin. Manning, C. 2006. The adaptation of early masonry churches in Ireland for use in later medieval times, 243-248. In: M. Meek (ed.), The modern traveller to our past: festschrift in honour of Ann Hamlin. DPK, Dublin. Marwick, H. 1952. Orkney. Edinburgh. Mason,T.H. 1938. The antiquarian remains of Inisheer, Aran, Co. Galway. Journal of the Royal Society of Antiquarians of Ireland, 68, 196-200. Masterton, R. 1998. Some lesser known ecclesiastical sites in Fore, Co. Westmeath. Riocht na Midhe, 9 (4), 40-48. Masterton, R. 2000. The church and the Anglo-Norman colonization of Ireland: a case study of the priory of Fore. Riocht na Midhe, 11, 58-70. Micklethwaite, J.T. 1896. Something about Saxon church building. Archaeological Journal, 53, 293-351. Micklethwaite, J.T. 1898. Some further notes on Saxon churches; being the Address at the opening of the Architectural Section of the Lancaster Meeting. Archaeological Journal, 55, 340-349. Minter, P., Potter, J.F. and Ryan, P. 2006. Roman and early medieval bricks and tiles: can they be distinguished? Essex Archaeology and History, 37, 95-102. Mytum, H. 1982. The location of early churches in northern County Clare, 351-361. In: S.M. Pearce (ed.), The early churches in Western Britain and 188

REFERENCES Ireland: studies presented to C.A. Ralegh Radford. BAR, British Series No.102, British Archaeological Reports, Oxford. Ní Ghabhláin, T. 1995. Church, parish and polity. The medieval diocese of Kilfenora, Ireland. Unpublished PhD thesis, University of California, Los Angeles, USA. Ní Ghrádaigh, J. 2008. Agha church: architectural iconography and building ambiguities, 53-83. In: T.McGrath, Carlow: history and society. Geography publications, Dublin. Ó Carragáin, T. 2005a. Habitual masonry styles and the local organisation of church building in early medieval Ireland. Proceedings of the Royal Irish Academy, 105C, 99-149. Ó Carragáin, T. 2005b. Regional variation in Irish PreRomanesque architecture. The Antiquaries Journal, 85, 23-56. O’ Keeffe, T. 1992. Romanesque architecture and sculpture at Ardmore, 73-104. In: W. Nolan and T.P. Power (eds), Waterford, history and society. Dublin. O’ Keeffe, T. 1998. Architectural traditions of the early medieval church in Munster, 112-124. In: M.A. Monk and J. Sheehan, Early medieval Munster: archaeology, history and society. Cork University Press, Cork. O’ Keeffe, T. 2003. Romanesque Ireland: architecture and ideology in the twelfth century. Four Courts Press, Dublin. O’ Keeffe, T. 2004. Ireland’s Round Towers: buildings, rituals and landscapes of the early Irish church. Tempus, Stroud. O’ Neill, M. 2002. The Medieval Parish churches in County Meath. Journal of the Royal Society of Antiquaries, 132, 1-56. O’ Reilly, F. 1997. Medieval Parish churches of northeast Clare. Sliabh Aughty (East Clare Heritage), 7, 31-37. Pålsson, H. and Edwards, P. 1981. Orkneyinga Saga: a history of the Earls of Orkney. Penguin, London. Parker, J.H. 1864. Ardmore (Notes on the architecture of Ireland – VII). The Gentleman’s Magazine, 134 (ii), 267-280. Parsons, D. (ed.) 1990. Stone: quarrying and building in England AD 43~1525. Phillimore, London. Parsons, D. 1991. Stone, 1-27. In: J.Blair, and N.Ramsey, English medieval industries: craftsmen, techniques, products. Hambledon Press, London. Pearson, A. 2003. The construction of the Saxon shore forts. British Archaeological Reports, British Series No. 349, BAR Publishing, Oxford. Pearson, A. and Potter, J.F. 2002. Church building fabrics on Romney Marsh and the marshland fringe: a geological perspective. Landscape History, 24, 89110. Peers, C.R. 1901. On Saxon churches of the St Pancras type. Archaeological Journal, 58. 402-434. Pentecost, A. 1993. British travertines: a review. Proceedings of the Geologists’ Association, 104, 2339. Petrie, G. 1845. The ecclesiastical architecture of Ireland, anterior to the Anglo-Norman invasion;

comprising an essay on the origin and uses of the round towers of Ireland. Vol. 1. Hodges and Smith, Dublin. Pevsner, N. and Williamson, E. 1994. The buildings of England: Buckinghamshire. Penguin, London. Potter, J.F. 1987. Geological traces of Saxon churches in the London Basin. Geology Today, SeptemberOctober, 164-168. Potter, J.F. 1988. Holes in the ground. The Environmentalist, 8 (1), 1-2. Potter, J.F. 1995. Let’s save our holes. Teaching Earth Science, 20 (2), 80. Potter, J.F. 1997. Green land misuse – a calamitous waste. The Environmentalist, 17 (4), 217-219. Potter, J.F. 1998. The distribution of silcretes in the churches of the London Basin. Proceedings of the Geologists’ Association, 109, 289-304. Potter, J.F. 1999. The geology of London Basin churches: the Palaeogene rocks. Tertiary Research, 19, 117-138. Potter, J.F. 2000a. The occurrence of travertine in the churches of the London Basin. Proceedings of the Geologists’ Association, 111, 55-70. Potter, J.F. 2000b. More about Tertiary churches: the London Clay septaria. Tertiary Research, 20, 59-63. Potter, J.F. 2000c. Protecting the last millennium. The Environmentalist, 20, 87-91. Potter, J.F. 2001a. The London basin’s gravel churches: indications of geology, medieval history and geographical distribution. Landscape History, 23, 526. Potter, J.F. 2001b. The occurrence of Roman brick and tile in churches of the London Basin. Britannia, 32, 119-142. Potter, J.F. 2002. The distribution of Bunter quartzites in the London Basin as portrayed by their abundance in church fabrics. Tertiary Research, 21, 155-170. Potter, J.F. 2003a. Continuing the debate on ferruginously-cemented gravel churches. Landscape History, 25, 79-82. Potter, J.F. 2003b. Geology of Essex churches – All Saints, Messing visited on a GA field meeting. Magazine of the Geologists’ Association, 2 (3), 6. Potter, J.F. 2004a. Viviparus limestone (‘Purbeck Marble’) – a key to financially well-endowed churches in the London Basin. Church Archaeology, 5-6, 80-91 and 2005, 7, 8 and 9, 192. Potter, J.F. 2004b. Suffolk’s Crag churches: with observations on other Crag building structures. Proceedings of the Suffolk Institute of Archaeology and History, 40, 399-413. Potter, J.F. 2005a. A geological review of some early Essex church quoins. Essex Archaeology and History, 36, 99-109. Potter, J.F. 2005b. No stone unturned – a re-assessment of Anglo-Saxon long-and-short quoins and associated structures. Archaeological Journal, 162, 177-214. Potter, J.F. 2005c. Ecclesiastical Geology: a return to Victorian field standards. Geoscientist, 15 (10), 4-7. Potter, J.F. 2005d. Field Meeting: Romney Marsh – its churches and geology, 22 May 2004. Proceedings of the Geologists’ Association, 116, 161-175. 189

PATTERNS IN STONEWORK: THE EARLY CHURCH IN BRITAIN AND IRELAND Potter, J.F. 2005e. Some observations on Orford Castle. Orford & District Local History Bulletin, 4, 4-8. Potter, J.F. 2005f. A geological review of some early Borderland churches. Transactions of the Shropshire Archaeological and Historical Society, 80, 1-16. Potter, J.F. 2006a. A geological review of some Hampshire Anglo-Saxon churches. Proceedings of the Hampshire Field Club and Archaeological Society, 61, 134-152. Potter, J.F. 2006b. Anglo-Saxon building techniques: quoins of twelve Kentish churches reviewed. Archaeologia Cantiana, 126, 185-218. Potter, J.F. 2006c. An analysis of ecclesiastical stone cut backs. Church Archaeology, 10, 57-80. Potter, J.F. 2006d. Stone emplacement in early Scottish churches: evidence of Early Christian craftsmanship. Proceedings of the Society of Antiquaries of Scotland, 136, 227-236. Potter, J.F. 2006e. A technique for distinguishing the textures of bricks and tiles. The Antiquaries Journal, 86, 326-330. Potter, J.F. 2007a. A day visiting churches: the geology of some West Surrey churches: 12th May 2007. Magazine of the Geologists’ Association, 6 (3), 14-15. Potter, J.F. 2007b. A review of some early West Sussex churches. Sussex Archaeological Collections, 145, 81-96. Potter, J.F. 2008a. The medieval town wall of Great Yarmouth, Norfolk: a geological perlustration. British Archaeological Reports, British Series No. 461, BAR Publishing, Oxford. Potter, J.F. 2008b. Looking at Jurassic churches – Saturday June 7 2008 (Part 1). Magazine of the Geologists’ Association, 7 (4), 16-17. Potter, J.F. 2008c. Early stone emplacement in three Scottish ecclesiastical monuments, Proceedings of the Society of Antiquaries of Scotland, 138, 205-222. Potter, J.F. 2008d. Decorative Anglo-Saxon wall fabrics: a pattern in stonework, Church Archaeology, 12 (in press). Potter, J.F. 2009a. Looking at Jurassic churches – Saturday June 7 2008 (Part 2). Magazine of the Geologists’ Association, 8 (1), 15-16. Potter, J.F. 2009b. A geological review of some early churches in the Northamptonshire area. Northamptonshire Archaeology (in press). Quin, C.W. and Wyndham-Quin, E.R.W. 1865. Memorials of Adare manor: with historical notices of Adare. Parker, Oxford. Reynolds, S.H. 1912. A geological excursion handbook for the Bristol district. J.W.Arrowsmith, Ltd, Bristol. Rickman, T. 1817. An attempt to discriminate the styles of English Architecture from The Conquest to the Reformation. Longman, London. Rickman, T. 1836. Further observations on the ecclesiastical architecture of France and England. Archaeologia, 26, 26-46. Ritchie, A. 1993. Viking Scotland. Batsford/Historic Scotland, London. Ritchie, A. and Ritchie, G. 1978. The ancient monuments of Orkney. HMSO, Edinburgh.

Roberts, E. 1974. Totternhoe Stone and flint in Hertfordshire churches. Medieval Archaeology, 18, 66-89. Rodwell, W. 1976. The archaeological investigation of Hadstock church, Essex: an interim report. The Archaeological Journal, 56, 55-71. Rodwell, W.J. 1998. Holy Trinity Church, Bradwelljuxta-Coggeshall: a survey of the fabric and appraisal of the Norman brickwork. Essex Archaeology and History, 29, 59-114. Rourke, G.D. and White Marshall, J. 2005. The drystone oratories of western Kerry, 103-124. In: J. White Marshall and C. Walsh, 2005, op. cit. Rowley, T. 1977. English Heritage book of Norman England. Batsford, London. RCAHMS (Royal Commission for Ancient and Historical Monuments for Scotland), 1920. Seventh Report: with inventory of monuments and constructions in the County of Dumfries. HMSO, Edinburgh. RCAHMS, 1928. Ninth Report: inventory of monuments and constructions in the Outer Hebrides, Skye and the small isles. HMSO, Edinburgh. RCAHMS, 1946. Twelfth Report: with an inventory of the ancient monuments of Orkney and Shetland. Edinburgh. Ruffino, J. 2000. The church and site of Temple Cronan, the Burran, County Clare. Unpublished MA. University College, Cork. Salter, A.E. 1905. On the superficial deposits of central and parts of southern England. Proceedings of the Geologists’ Association, 19 (1), 1-56. Salzman, L.F. 1967. Building in England down to 1540. Clarendon Press, Oxford. Sellwood, B.W. 1984. The rock types represented in the town wall of Silchester, 224-231. In: Silchester. Excavations on the defences 1974-80. Britannia Monograph Series, 5, Society for the Promotion of Roman Studies, London. Sevastopulo, G.D. 1981. Lower Carboniferous, 147-171. In: C.H.Holland (ed.), A geology of Ireland. Scottish Academic Press, Edindurgh. Sharkey, O. 1999. Fore and its ancient buildings. Magpie Publications, Westmeath. Simpson, A. 1963. The stratigraphy and tectonics of the Manx Slate Series, Isle of Man. Quarterly Journal of the Geological Society of London, 119, 367-400. Smith, J.T. 1972. Ardmore Cathedral. Journal of the Royal Society of Antiquaries of Ireland, 102, 1-13. Spicer, A. 2000. “Defyle not Christ’s kirk with your carrion”: burial and the development of burial aisles in post-Reformation Scotland, 149-169. In: B.Gordan and P. Marshall (eds), The place of the dead. Death and remembrance in late medieval and early modern Europe. Cambridge University Press, Cambridge. Stalley, R. 2000. Irish round towers. Town House and Country House Ltd, Dublin. Stillman, C.J. 1981. Caledonian igneous activity, 83-106. In: C.H.Holland (ed.), A geology of Ireland. Scottish Academic Press, Edinburgh. Stocker, D. and Everson, P. 1990. Rubbish recycled: a study of the re-use of stone in Lincolnshire, 83-101. 190

REFERENCES In: D.Parsons (ed.), Stone: quarrying and building in England AD 43~1525. Phillimore, London. Stocker, D. and Everson, P. 2006. Summoning St Michael: Early Romanesque towers in Lincolnshire. Oxbow Books, Oxford. Sutherland, D.S. 1990. Burnt stone in a Saxon church and its implications, 102-113. In: D. Parsons (ed.), Stone: quarrying and building in England AD 43~1525. Phillimore, London. Sutherland, D.S. 2000. Ecclesiastical geology, 292-295. In: P.L. Hancock and B.J. Skinner (eds), The Oxford companion to the earth. Oxford University Press, Oxford. Sutherland, D.S. 2003. History in stone – Brixworth Saxon church, 112-118. In: D.S. Sutherland, Northamptonshire Stone. The Dovecote Press, Dorset. Sutherland, D.S. and Parsons, D. 1984. The petrological contribution to the survey of All Saints’ Church, Brixworth, Northamptonshire: an interim study. Journal of the British Archaeological Association, 137, 45-64. Swan, D.L. 1994. Early monastic sites, 136-139. In: M Ryan (ed.), Irish archaeology illustrated. Country House, Dublin. Tatton-Brown, T. 1980. The use of Quarr Stone in London and East Kent. Medieval Archaeology, 24, 213-215. Tatton-Brown, T. 2001. The quarrying and distribution of Reigate Stone in the Middle Ages. Medieval Archaeology, 45, 189-201. Taylor, H.M. 1961. English architecture in the time of Bede. Rector of Jarrow, Jarrow. Taylor, H.M. 1969. The special role of Kentish churches in the development of Pre-Norman (Anglo-Saxon) architecture. Archaeological Journal, 126, 192-198. Taylor, H.M. 1978. Anglo-Saxon architecture. Vol. III. Cambridge University Press, Cambridge. Taylor, H.M. and Taylor, J. 1965. Anglo-Saxon architecture. 2 vols. Cambribge University Press, Cambridge. The Corpus of Romanesque sculpture in Britain and Ireland. (www. crsbi.ac.uk). Thomas, C. 1971. The early Christian Archaeology of North Britain. Oxford University Press, London. Waddell, J. 1972-1973. An archaeological survey of Temple Brecan, Aran. Journal of the Galway Archaeological and Historical Society, 33, 7-27. Waddell, J. and Clyne, M. 1995. M.V. Duignan’s excavations at Kiltiernan, Co. Galway, 1950-1953. Journal of the Galway Archaeological and Historical Society, 47, 149-203. Waller, J.A. 1846. Notes on Anglo-Saxon masonry. Journal of the British Archaeological, 1, 117-120. Waterman, D.M. 1967. The early Christian churches and cemetery at Derry, Co. Down. Ulster Journal of Archaeology, 3rd Series, 30, 53-75. Waterman, D.M. 1970. Somersetshire and other foreign building stone in medieval Ireland, c. 1175-1400. Ulster Journal of Archaeology, 3rd Series, 33, 63-75. Waterman, D. and Hamlin, A. 1976. Banagher Church, Co. Derry. Ulster Journal of Archaeology, 39, 25-41.

Wentworth, C.K. 1922. A scale of grade and class terms for clastic sediments. The Journal of Geology, 30, 377-392. Westropp, T.J. 1895a. Aran Islands. Journal of the Royal Society of Antiquaries of Ireland, 25, 250-278. Westropp, T.J. 1895b. Barony of Burren, Co. Clare. Journal of the Royal Society of Antiquaries of Ireland, 25, 279-284. Westropp, T.J. 1900-1902. The churches of County Clare, and the origin of the ecclesiastical divisions in that county. Proceedings of the Royal Irish Academy, Third Series, 6, 100-180. Westropp, T.J. 1903. Notes on the antiquities of Ardmore. Journal of the Royal Society of Antiquaries of Ireland, 33, 353-380. Westropp, T.J. 1905. A survey of the ancient churches in the County of Limerick. Proceedings of the Royal Irish Academy (Section C), 25 (8), 327-480. Westropp, T.J. 1907. A description of the ancient buildings and crosses of Clonmacnois, King’s County. Journal of the Royal Society of Antiquaries of Ireland, 37, 277-306. Westropp, T.J. 1909. Notes on the antiquities around Kilfenora and Lehinch, Co. Clare. Journal of the North Munster Archaeological Society, 1 (1), 14-29. Westropp, T.J. 1919. Appendix I. – Notes on several forts in Dunkellin and other parts of southern Co. Galway. Journal of the Royal Society of Antiquaries of Ireland, 49, 167-186. White Marshall, J. and Walsh, C. 1998. Illaunloughan, Co. Kerry: an island hermitage, 102-111. In: M.A. Monk and J. Sheehan, 1998. Early Medieval Munster: Archaeology, History and Society. Cork University Press, Cork. White Marshall, J. and Walsh, C. 2005. Illaunloughan Island: an early medieval monastery in County Kerry. Bray, Co. Wicklow: Wordwell Ltd. Whittow, J.B. 1974. Geology and scenery in Ireland. Penguin, London. Woods, J. 1907. Annals of Westmeath ancient and modern. Sealy, Bryers and Walker, Dublin. Worssam, B.C, and Tatton-Brown, T. 1990. The stone of the Reculver columns and the Reculver cross, 51-69. In: D. Parsons (ed.), Stone: quarrying and building in England AD 43~1525. Phillimore, London. Worssam, B.C. and Tatton-Brown, T. 1993. Kentish Rag and other Kent building stones. Archaeologia Cantiana, 112, 93-125. Zarnecki, G. 1951. English Romanesque Sculpture 10661140. London. Zarnecki, G. 1953. Later English Romanesque Sculpture 1140-1210. London. Zarnecki, G., Holt, J. and Holland, T. (eds) 1984. English Romanesque Arts, 1066-1200. London.

191