A study of Dundonnell Castle, Co. Roscommon

A STUDY OF DUNDONNELL CASTLE, CO. ROSCOMMON

Daniel Curley, B.A.

In partial fulfilment of the MA in Medieval Studies

National University of Ireland, Galway

July 2011

Dr. Stefan Bergh, Archaeology Department Head

Dr. Kieran O‟Conor, Supervisor

ii

Table of Contents

Table of Figures .....................................................................................................................iv

Table of Plates ......................................................................................................................... v

Abstract ....................................................................................................................................vi

Acknowledgements ............................................................................................................ vii

Chapter 1: Aims, Methods and Sources .......................................................................... 1 1.0 – Introduction ........................................................................................................................... 1 1.1 - The aims of this thesis ......................................................................................................... 4 1.2 – Methodology ........................................................................................................................... 5 1.3- Previous work on Dundonnell .......................................................................................... 7 1.4 – Sources used in this thesis .............................................................................................. 10

1.4a - Documentary Sources: .............................................................................................................. 10 1.4b – Cartographic Sources ............................................................................................................... 13

1.5 – Conclusions ........................................................................................................................... 15

Chapter 2-The Historical Background ........................................................................ 16 2.0 – Introduction to the Chapter ............................................................................................ 16 2.1 - Early Medieval Ireland (c. 500AD – c. 1100AD) ....................................................... 16 2.2 - High Medieval Ireland (c. 1100 – c. 1380) .................................................................. 25 2.3 - Late Medieval Ireland (c. 1380 – c. 1650) .................................................................. 34 2.4 - Taughmaconnell parish .................................................................................................... 39 2.5 - Historical references to Onagh....................................................................................... 41

Chapter 3 – Physical description of the site .............................................................. 43 3.0 – Introduction ......................................................................................................................... 43 3.1 – Siting ....................................................................................................................................... 43 3.2 – Description of the Site (Fig. 8) ....................................................................................... 44

3.2a – The Earthwork ............................................................................................................................ 44 3.2b – The stronghouse ........................................................................................................................ 47

3.3 – Conclusions ........................................................................................................................... 57

Chapter 4 - The Siting Chapter ....................................................................................... 60 4.0 - Introduction .......................................................................................................................... 60 4.1 - Ringfort Siting ...................................................................................................................... 60 4.3 - Defensibility of the site ..................................................................................................... 70 4.4 - Conclusions ........................................................................................................................... 72

Chapter 5 - Phase 1 at Dundonnell: the postulated Ringfort phase .................. 74 5.0 - Introduction .......................................................................................................................... 74 5.1 - The postulated ringfort at Dundonnell ....................................................................... 74 5.2 - The ringfort in Ireland ...................................................................................................... 77

5.2a - Number & Distribution ............................................................................................................. 77 5.2b - Morphology ................................................................................................................................... 77 5.2c - Function .......................................................................................................................................... 78 5.2d - Dating .............................................................................................................................................. 80

5.3 - Conclusions ........................................................................................................................... 81

Chapter 6 - Phase 2 at Dundonnell: The postulated Ringwork .......................... 82 6.0 - Introduction .......................................................................................................................... 82

iii

6.1 - The postulated ringwork at Dundonnell .................................................................... 82 6.2 -The ringwork in Ireland .................................................................................................... 85

6.2a - Number and Distribution ........................................................................................................ 85 6.2b - Morphology ................................................................................................................................... 85 6.2c - Defences .......................................................................................................................................... 87 6.2d - Function ......................................................................................................................................... 87 6.2e - Dating............................................................................................................................................... 89

6.3 - Conclusions ........................................................................................................................... 89

Chapter 7 - The Third Phase at Dundonnell: The Stronghouse .......................... 91 7.0 - Introduction .......................................................................................................................... 91 7.1 - The stronghouse at Dundonnell .................................................................................... 91 7.2 – Stronghouse or Fortified House? .................................................................................. 92 7.3 - The outer defences at Dundonnell ................................................................................ 93 7.4 - Stronghouses in Ireland.................................................................................................... 93 7.5 - Other stronghouses in Roscommon ............................................................................. 96 7.6 - Conclusions ........................................................................................................................... 97

Chapter 8- Conclusions ..................................................................................................... 99 8.0 - Introduction .......................................................................................................................... 99 8.1 - Main conclusions of the thesis ........................................................................................ 99 8.2 - Future work ....................................................................................................................... 100

Primary Source Bibliography ...................................................................................... 101

Secondary Source Bibliography ................................................................................. 102

iv

Table of Figures

Figure 1: Location of Dundonnell Castle within Co. Roscommon ...................................... 1 Figure 2: Graham's sketch plan and cross-section (1988) ................................................... 9 Figure 3: Craig's ground floor sketch plan of stronghouse (1989) ................................... 9 Figure 4: Taughmaconnell in the Stafford Survey Map for the Barony of Athlone

(1636) .......................................................................................................................................... 14 Figure 5: Six Inch 1st edition Ordnance Survey Map of Dundonnell (1829) ................ 15 Figure 6: Territory of the kingdom of Uí Maine (expansion & decline taken into

account) ....................................................................................................................................... 20 Figure 7: Map of the King's Cantreds .......................................................................................... 31 Figure 8: Dundonnell 1829 Ordnance Survey Map (Figure 5 repeated) ....................... 44 Figure 9: Total Station of the earthwork, with the stronghouse situated in the north-

western quadrant. ................................................................................................................... 46 Figure 10: Reconstruction of what may have existed atop the earthwork at

Dundonnell c. mid 13th-century, courtesy of Carrie O'Malley (2011) ................. 47 Figure 11: Ground Floor plan of Stronghouse ......................................................................... 52 Figure 12: First floor plan of the stronghouse ........................................................................ 55 Figure 13: Reconstruction of the south facing walls of the stronghouse at

Dondonnell, courtesy of Carrie O'Malley (2011) .......................................................... 58 Figure 14: Reconstruction of the north walls of the stronghouse at Dundonnell,

courtest of Carrie O'Malley (2011) .................................................................................... 59 Figure 15: Location of Dundonnell with regard to the major medieval routeways of

the region .................................................................................................................................... 65 Figure 16: Ordnance Survey Map containing the extents of Dundonnell and Onagh

townlands ................................................................................................................................... 77

v

Table of Plates

Plate 1: Dundonnell Castle ............................................................................................................... 2 Plate 2: North-eastern wall, largely broken out ..................................................................... 49 Plate 3: Ground floor fireplace and kitchen location on south-eastern gable wall ... 50 Plate 4: Gunloop on south-western wall (interior view) .................................................... 50 Plate 5: Gunloop on south-western wall (exterior wall) ..................................................... 50 Plate 6: Joist holes and corbel on south-western wall interior ........................................ 52 Plate 7: Windows on south-western wall exterior ................................................................ 53 Plate 8: North-western gable wall interior .............................................................................. 54 Plate 9: Evidence of machicolation on south-eastern gable wall ..................................... 57 Plate 10: Diamond-shaped chimneystacks .............................................................................. 57

vi

Abstract

Dundonnell Castle will be explored through a multi-disciplinary approach. This study

aims to provide a greater understanding of a simultaneously neglected but oft

mentioned potentially multi-period site that has been given many roles in previous

scholarly work. This is without ever having been studied properly and fully in its own

right. This thesis combines an archaeological analysis of the earthworks and masonry

building at Dundonnell with a large corpus of historical material that pertains to the

immediate and surrounding area over an extended period of time. More specifically, the

thesis will attempt to recreate the importance of the three possible phases at the site, via

the use of historical, cartographical, and archaeological evidence to demonstrate the

site‟s dates of importance, function and role in the wider areas, and to try and pinpoint

the site on the historical plane. The study of the castle type, known as the stronghouse,

will also be examined in relation to Dundonnell, in an attempt to further the knowledge

of this understudied monument type.

vii

Acknowledgements

Firstly, I would like to thank the Archaeology, History and Classics departments at

NUI, Galway for their inspiration and the dedication that they bring to their respective

disciplines. To this I would like to also show my appreciation for the staff of the said

departments and others, who helped me throughout my MA. Their enthusiasm was a

great motivation for my own work and I hope I showed my appreciation for their

efforts.

This leads me to Dr Kieran O‟Conor. Thank you for your direction and stalwart

encouragement and backing throughout the process of this study, along with providing

the spark for the thesis that accompanies this note. It was with your passion and love of

your work that inspired my own humble effort, and my hope is that it goes some of the

way to embodying the spirit of the MA.

I would like to thank also the immeasurable help provided by Joseph Fenwick and Rory

Sherlock. Joseph‟s hours of dedicated teaching in the mediums of digital surveying

were an invaluable addition to my work, along with being intensely interesting and

thoroughly enjoyable. The work done by Rory Sherlock provided a fresh pair of eyes

on the physical remains of Dundonnell, along with helping to plan said remains in a

manner best to analyse the site and, more specifically, its masonry phase. Thank you

both for your time, assistance and knowledge.

Thanks to Dr Mark Stansbury for your insightfulness and knowledge, and most of all

for your approachability and belief in my cause throughout the two years of the MA.

Our conversations were so often reassuring and your determination to further multi-

disciplinary studies has proved to be a major influence on my thought process, and for

that enlightenment I cannot possibly be able to adequately show my appreciation.

I would like to extend my gratitude to Kevin Flynn, the owner of the lands that

surround Dundonnell Castle for his permission to begin a study of this site and for his

wholehearted good wishes for my endeavours.

I would like to thank Josephine, Danny, Aishling and Niall Curley for their hard work

and grit in attempting to tidy up the swathes of overgrowth that hid Dundonnell, and

their patience with my attempts to master the Digital Survey. I would also like to thank

Denis Judge, for his generosity in providing me with the tools and equipment necessary

to help the work run smoothly and safely. I would also like to thank Carrie and Danielle

O‟Malley. Carrie for your immense artistic talents that served to illustrate my study,

and to Danielle for being you, and for keeping me sane and grounded when the stresses

of the research were starting to show! To all the above, I cannot even begin to show my

eternal indebtedness to your talents of listening and patience.

I wish to extend thanks to my colleagues in my MA, Mairead Keane, Sean Sullivan and

Kenneth Coyne, for being the like-minded individuals that led to a great working

relationship and constant help, but more importantly to being great and thoughtful

friends. Truer crusaders of the cause do not exist!

1

MA in Medieval Studies thesis (Archaeology)

A Study of Dundonnell Castle, Co. Roscommon

Chapter 1: Aims, Methods and Sources

1.0 – Introduction

Dundonnell Castle consists of a three-storey late 16th

- or early-to-mid 17th

-century

stronghouse located with an oval, bivallate ditched and banked enclosure (RMP NO.

RO047002, 047002) that lies today in undulating pasture in modern south Roscommon

within the bounds of the parish of Taughmaconnell (Teach Mhic Conaill – the House of

the Sons of Connell) (Fig. 1; Pl. 1).

Figure 1: Location of Dundonnell Castle within Co. Roscommon

2

Plate 1: Dundonnell Castle

The site lies 2km northwards from my home place and I have been interested in it all

my life. However, little detailed research has been carried out on the site, which is

surprising given the fact that it is quite well preserved. It is true that a few scholars have

noted the site in various publications.1 Some scholars have suggested that the shape of

the oval bivallate enclosure and the placename may be an indication that Dundonnell

was originally a reused pre-Norman ringfort, an Anglo-Norman ringwork castle or

both.2 In other words, the general consensus is that the stronghouse was built within a

1 Graham, B., “Medieval Timber and Earthwork Fortifications in Western Ireland”, Medieval

Archaeology (1988), pp. 110-129; Graham, B., “Medieval Settlement in County Roscommon”,

Proceedings of the Royal Irish Academy. Section C: Archaeology, Celtic Studies, History, Linguistics,

Literature, Vol. 88C (1988), pp. 19-38; Craig, M., The Architecture of Ireland from the earliest times to

1880, (London & Dublin: 1989); O‟Keefe, T., “The Fortifications of Western Ireland, AD 1100-1300,

and Their Interpretation”, Journal of the Galway Archaeological and Historical Society, Vol. 50 (1998),

pp. 184-200; Sweetman, D., The Medieval Castles of Ireland, (Cork: 1999); Doran, L., “Medieval

Communication Routes through Longford and Roscommon and Their Associated Settlements”,

Proceedings of the Royal Irish Academy. Section C: Archaeology, Celtic Studies, History, Linguistics,

Literature, Vol. 104C, No. 3 (2004), pp. 57-80; O‟Conor, K., “English settlement and change in

Roscommon during the late sixteenth and seventeenth centuries”, in Horning, A., Ó Baoill, R., Donnelly,

C., Logue, P.,(eds.) The Post Medieval Archaeology of Ireland 1550-1850, (2007), pp. 189-203 2 Graham, “Medieval Timber and Earthwork Fortifications in Western Ireland”, pgs. 122-123; Graham,

“Medieval Settlement in County Roscommon”, pgs. 28-29

3

pre-existing earthwork, which was then itself re-used as peripheral defences around the

latter building.

On one level, the fact that the masonry stronghouse seems to have been built within an

earlier monument of general medieval date means that the site of Dundonnell qualifies

as a suitable candidate for detailed study as part of NUI, Galway‟s MA in Medieval

Studies. In this respect, a legalistic, possibly Anglo-centric view might be that the

stronghouse at the site is not medieval in date and so this particular aspect of the

monument should not be studied in depth. However, certain Irish archaeologists, such

as Martin Jope, Tom McNeill and Kieran O‟Conor, have argued strongly that the

medieval period in Ireland ended in archaeological terms around 1650. This is argued

to have been particularly true of the Midlands and western half of the country.3 Duffy,

Edwards and FitzPatrick, in the 2001 book Gaelic Ireland c. 1260 –c. 1650: Land,

Lordship and Settlement state that the medieval period ends in Ireland around 1600.

The post medieval period is seen as beginning sometime around the latter year.4 It will

be argued below that the best time to see the construction of the stronghouse is

sometime in the last three decades of the 16th

-century – well within the time limits of

the medieval period in Ireland as set by scholars like McNeill and O‟Conor and even

Duffy, Edwards and FitzPatrick. Therefore, it is argued here that it is justified to study

the stronghouse in this thesis, as its construction falls within what scholars believe to be

the medieval, in particular the late medieval, period.

3 Jope, E. M., Jope, H. M., & Johnson, E. A., The Archaeological Survey of Co. Down, (Belfast: 1966);

McNeill, T. E., Castles in Ireland: Feudal Power in a Gaelic World, (London and New York: 1997) pgs.

228-299; O‟Conor, K., The Archaeology of Medieval Rural Settlement in Ireland, (Dublin: 1998) p. xi 4 Duffy, P. J., Edwards, D., & FitzPatrick, E., Gaelic Ireland c. 1250-1650: Land, Lordship and

Settlement, (Four Courts Press, Dublin: 2001), p. 17

4

The aim of the rest of this chapter is to outline the exact aims of this thesis, along with

the methods and sources used within it.

1.1 - The aims of this thesis

One of the aims of the thesis is to gain a better understanding of the way the site at

Dundonnell developed through time. Can we figure out what was the function (s) of the

site during different periods of its usage?

Very little work has been carried out on the stronghouse as a castle type. The

stronghouse belongs to the late 16th

and early-to-mid 17th

-centuries, but the information

known about castles from this period in Ireland usually begins and ends with the

detailed discussion of its more illustrious and elegantly designed cousins, the tower

house and fortified house. The latter type of castle balances Renaissance-influenced

aesthetics and defence in a way that the stronghouse cannot compare, and as a result,

their study has been largely neglected. As such, a major aim of this thesis is use the site

at Dundonnell as a means of learning more about the stronghouse in terms of its

architecture, function, construction, dating, use of internal space and defences.

No proper plan or description exists of the site at Dundonnell. Therefore, the last major

aim of the thesis is to carry out a Total Station survey of the site that will provide a

modern plan. A detailed description of the site will also be provided, which can be used

for future comparative studies elsewhere. It is clear that one of the weaknesses of Irish

castle studies (and, indeed, medieval studies in this country) is that most castles in

Ireland have not been properly planned and described. This is in direct contrast to the

5

situation in England, Wales and Scotland. It has been stated that for Irish castle studies

to progress, more plans and detailed, scientific descriptions of castles are needed.5 A

good plan and description of Dundonnell will be a valuable addition to medieval castle

studies in Ireland and will provide invaluable, trustworthy comparative data for anyone

interested in studying stronghouses in the future.

1.2 – Methodology

The methodology used in this thesis is based upon the interdisciplinary spirit that is

central to the ideas of the MA in Medieval Studies at NUI, Galway, and a method that

in this case combines historical documents, historical commentaries, archaeological

survey, analysis and GPS technology, cartographical studies and antiquarian sources in

order to achieve the aims provided above. This I feel is the best way of throwing light

on the monument at Dundonnell. It is an inclusive approach that leaves no avenue

unexplored and thus takes into account every possible situation before coming to any

conclusion.

This approach may seem logical but it is not always taken. For example, Sweetman‟s

1999 book The Castles of Ireland takes a purely archaeological approach. The historical

context of the castles analysed is not discussed and this means that the functions and

exact dates of these monuments is not fully understood by the latter author.6 In contrast,

it has recently been stated by the American settlement historian and archaeologist

Thomas Finan that “the exciting and innovative work in Irish medieval studies is taking

place in the margins between disciplines, where methodological and disciplinary

5 O‟Conor, K., “Castle Studies in Ireland”, Chateau Gaillard 23 (2008), pp. 329-39, pgs. 329-31

6 Sweetman, The Medieval Castles of Ireland

6

preoccupations break down and evidence can be re-analysed from a new perspective.”7

It is just such an interdisciplinary approach that I want to take at Dundonnell.

For the purposes of this thesis, a detailed archaeological survey of Dundonnell Castle,

the masonry monument and the earthwork, will take place, in order to record and

interpret the physical remains of the site, thus providing us with another form of

information with which to gain a better understanding of the monument and its place in

history. To do this, the masonry castle itself must be measured and planned out, with

specific digital tools to be applied to the earthwork. This comes in the form of digital

surveying - with the Trimble GPS, and the Total Station. The information gathered

from these devices can be collected, stored, and analysed on ArchGIS and Geosite

Office, programs that specialise in map production and analysis, to show how the site

was constructed, and to help understand and highlight any aspects of the monument that

may have gone unnoticed to the naked eye. These tools, therefore, add vital information

to this thesis, especially considering that the site will, unfortunately, not be excavated

for the purposes of this study.

The Trimble GPS (Global Positioning System) works by using satellites to pinpoint the

device‟s position on the landscape up to a centimetre of accuracy, and thus, the holder

can plot the device‟s position at regular intervals to record a feature on the landscape on

the Irish National Grid. With this, the archaeologist can create a map detailing all the

various features that make up a given site and its hinterland, providing important

distance and gradient information, drawing and plotting vital features, all via the

7 Finan, T., “Introduction: Moylurg and Lough Ce in the Later Middle Ages” in Finan, T., (ed.) Medieval

Lough Ce – History, Archaeology and Landscape, (Dublin: 2010), pp. 11-14, p. 11

7

National Grid. Its applications in regard to the site at hand would centre around aiding

the use of the Total Station in providing OS co-ordinates from which to work from, a

vital component of the digital process, without which the digital survey could never be

applied to the National Grid.

The Nikon DTM-322 Total Station is another vital tool in the surveying of

archaeological sites, through its utilisation of the recording of a series of X, Y & Z co-

ordinates on a plane that can again be placed in the co-ordinates of the Irish National

Grid. The Total Station itself uses a laser which it reflects off a prism situated on a

detail pole at whatever point has been chosen for recording. When each co-ordinate is

recorded and downloaded to the map producing programs of Geosite Office or

ArchGIS, they plot the landscape and can be manipulated to create anything from a line

drawing of the site, which is our primary concern, to 3D maps of the site, with the

crucial benefit of being able to pick up features that would be invisible to the naked

eye. Its applications here will be important to show the scale and details of the

earthwork, identifying specifics such as the entrance and the magnitude of the ditch and

bank systems. It is with the analysis of these details that we can garner an

understanding of how important this site may have been during its periods of

occupation, along with why it was positioned where it was and the benefits inherent to

same.

1.3- Previous work on Dundonnell

The first modern scholarly discussion of Dundonnell Castle, albeit brief, surfaces with

the work of Brian J. Graham in his two 1988 articles “Medieval Timber and Earthwork

8

Fortifications in Western Ireland” published in the journal Medieval Archaeology8 and

“Medieval Settlement in County Roscommon” published in the Proceedings of the

Royal Irish Academy.9 Graham provides a short description and some information on

the site and is mainly concerned with its ringwork/ringfort phase, inserting both a rough

sketch plan and a cross-section of the earthwork in his second article (Fig 2).10

Maurice

Craig briefly mentions and includes an almost miniscule sketch plan of the stronghouse

in his 1989 book The Architecture of Ireland from the Earliest Times to 1880. He dates

this building to the first half of the 17th

century (Fig. 3).11

8 Graham, “Medieval Timber and Earthwork Fortifications in Western Ireland”, pgs. 122-123

9 Graham, “Medieval Settlement in County Roscommon”, pgs. 28-29

10 Graham, “Medieval Settlement in County Roscommon”, Fig. 3, p. 28

11 Craig, M., The Architecture of Ireland from the earliest times to 1880, pgs. 131 & 134

9

Figure 2: Graham's sketch plan and cross-section

(1988)

Figure 3: Craig's ground floor sketch plan of

stronghouse (1989)

Tadhg O‟Keeffe briefly discusses Dundonnell Castle also in his 1998 article “The

Fortifications of Western Ireland, AD 1100-1300, and Their Interpretation”, published

in The Journal of the Galway Archaeological and Historical Society.He believes that

the stronghouse was built within a pre-existing, unoccupied ringfort and does not

believe that the latter earthwork was re-used as a ringwork castle in the 13th

-century –

meaning that he sees only two phases at the site.12

David Sweetman also briefly

mentions Dundonnell on the basis of its stronghouse phase in his all-encompassing

1999 work The Medieval Castles of Ireland.13

However, it is again unfortunately

mentioned merely as one of a number of stronghouses that exist in Roscommon and is

not treated to any great extent on its own unique characteristics. Linda Doran‟s study of

medieval communication routes and roadways through the midlands also makes

mention of Dundonnell and notes its location in close proximity to both the Slíghe

12

O‟Keefe, T., “The Fortifications of Western Ireland, AD 1100-1300, and Their Interpretation”, p. 190

and note 54, p. 198 13

Sweetman, D., The Medieval Castles of Ireland, pgs. 197-198

10

Mhór (a major roadway of the time) and to an important secondary road perpendicular

to the Slíghe Mhór, which Dundonnell lies beside.14

Kieran O‟Conor discusses the site

in regard to its masonry stronghouse phase in his 2007 article on English settlement in

Roscommon during the late 16th

and early 17th

-centuries, and raises interesting

questions as to the history and the purpose of the site. He argues that there was no

Anglo-Norman ringwork phase at the site – believing that the site was originally a

ringfort refortified in Tudor and Jacobean times.15

Compilers at the Archaeological

Survey of Ireland visited the Dundonnell in the mid 1990s and classified the site as an

Anglo-Norman ringwork castle within which was situated a late 16th

or early-to-mid

17th

-century fortified house.16

In summary, it is clear that Dundonnell has been recognized as an important site by a

number of scholars but little detailed work has been carried out on it. No proper plan or

comprehensive description of the site has been published – a situation replicated right

across the country, as noted.

1.4 – Sources used in this thesis

1.4a - Documentary Sources:

Documentary sources are a vital addition to any archaeological study and can be used

by archaeologists in a variety of ways.17

In the case of this thesis, documentary

evidence will seek to play an even more important role, due to the author‟s own

14

Doran, L., “Medieval Communication Routes through Longford and Roscommon and Their

Associated Settlements”, p. 76 15

O‟Conor, K., “English settlement and change in Roscommon during the late sixteenth and seventeenth

centuries”, pgs. 191-192 16

Archaeological Survey of Ireland, 2010, National Monuments Service [online], available:

http://webgis.archaeology.ie/NationalMonuments/flexviewer/[accessed 9th June 2011] 17

Barry, T., The Archaeology of Medieval Ireland, (London & New York: 1987), pgs. 3-10

11

primary training in the historical discipline. Evidence from these historical sources can

help date and phase sites such as Dundonnell. They can provide information about the

occupants of any given site and inform us about the general political, economic and

social context within which the inhabitants of places like Dundonnell operated. They

can also help in understanding how a site looked when it was in use. For example, O‟

Conor has successfully used surviving late 13th

and early 14th

-century manorial extents

to reconstruct the appearance of certain motte castles in Leinster around c. 1300.18

Finally, we must not forget to analyse the language in use in the contemporary written

sources, as what is said, how it is phrased and what is potentially omitted can colour the

situation to an even greater degree, allowing us to understand as much as we can about

the topic at hand. The historical sources that will come into play in this thesis include

the native Irish annals and edited colonial sources emanating from the Dublin

government such as the four volumes of the Calendar of Documents Relating to

Ireland, 1171-130719

and the later twenty-three volumes of the Calendar of State

Papers, Ireland.20

Both colonial and native sources must be employed with caution, but

indeed it is this inherent bias in the texts that is often of most interest.

The Gaelic annals are an essential source of information for areas that saw little

intensive English settlement, such as modern Roscommon and much of the Midlands

and western Ireland. The entries within these annals are laconic in nature but are full of

references to ecclesiastical and secular settlements in the latter regions. They do

provide a certain amount of information about the economy and society of the more

18

O‟Conor, K., The Archaeology of Medieval Rural Settlement in Ireland, (Dublin: 1998), pgs. 28-33 19

Calendar of documents relating to Ireland, 1252-1284, Sweetman, H. S. (ed.), (Nendeln, 1974) 20

Calendar of State Papers relating to Ireland of the reigns of Henry VIII, Edward VI, and Elizabeth,

1509-[1603]. Hamilton, H. C. (ed.). P.R.O. (London: 1920-31)

12

Gaelic-dominated parts of the country, such as the Dundonnell area, after 1169 and, of

course, before that date as well.21

The entries in these annals are not recorded in any

given order on an annalistic year and they are simple notes recording an event that has

occurred in that year. This is a problem that is endemic of annals, but their usefulness

cannot be understated for the recording and comparison of events, and often for

providing events mentioned in literature and in longer narratives with a date, if it hadn‟t

been mentioned in the text itself. The annals of most interest to this study as that of the

Annals of Clonmacnoise edited by Rev. D. Murphy and published in 189622

, the Annals

of Connacht, edited by A. D. Freeman and published in 194423

, and The Annals of the

Four Masters, the four volumes of which were compiled and edited by John

O‟Donovan between 1848 and 1854.24

The Annals of Clonmacnoise, the most

important to my research,deals with the period from the mythical Biblical creation of

man until the year 1408, with a few entries missing. It gives the history of the island of

Ireland and the area surrounding Clonmacnoise in particular, which is very useful,

given its proximity to Dundonnell itself.

The colonial sources, particularly those emanating from the central government in

Dublin, are far more detailed in comparison to the native sources, both in terms of

numbers surviving and information.25

The five volumes of the Calendar of Documents

Relating to Ireland, 1171-1307 were used in this thesis and contain a multitude of

21

Barry, T., The Archaeology of Medieval Ireland, p. 9 22

The Annals of Clonmacnoise: Being annals of Ireland from the earliest period to A.D. 1408, Murphy,

Rev. D. (ed.), (Dublin: 1896, facsimile reprint 1993) 23

Annála Connacht: the Annals of Connacht, AD. 1224-1544, Freeman, A.D. (ed.), (Dublin: 1944; reprint

1977) 24

Annals of the kingdom of Ireland by the Four Masters: from the earliest period to 1171, O‟Donovan, J.,

(ed.), (Hodges and Smith, Dublin: 1854) 25

Barry, T., The Archaeology of Medieval Ireland, pgs. 3-9

13

references to Anglo-Norman castles and manorial centres.26

The 23 volumes of the

Calendar of State Papers, Ireland, edited by various authors in the 19th

and early 20th

century, are of importance to our general understanding of the stronghouse phase at

Dundonnell. These volumes contain a correspondence between English administrators

in Ireland and their counterparts and superiors at the seat of the royal government and

court in London.27

In addition to these sources, broader contemporary sources will be

consulted to understand the site and its hinterland in the years of its occupation, as

specifically in the cases of its ringwork and stronghouse phases, due to them coinciding

with periods of considerable transition and political and military activity in the region.

1.4b – Cartographic Sources

Late 16th

- and early-to-mid 17th

-century maps linked to the Tudor and Stuart reconquest

of Ireland are dotted with depictions of castles, artillery forts, abbeys, churches,

routeways and secular urban and rural settlements. These maps can be of great value to

archaeologists attempting to reconstruct the settlement patterns and landscape of late

medieval Ireland. These maps can also help date sites under study.28

The English-

controlled government in Dublin in the 17th-century ordered a number of surveys and

inquisitions. Much of these maps were drawn up to facilitate the large-scale

confiscation and redistribution of land to loyal English and Scots Protestant settlers.

One of these surveys was the Stafford Inquisition of Connacht in 1636. These included

26

Calendar of documents relating to Ireland, 1252-1284, & Barry, T., The Archaeology of Medieval

Ireland, p. 4 27

Calendar of State Papers relating to Ireland of the reigns of Henry VIII, Edward VI, and Elizabeth,

1509-[1603] 28

Prunty, J., Maps and Map Making in Local History, (Dublin: 2004)

14

maps, including one of the Barony of Athlone, within which Dundonnell townland is

situated (Fig.4).29

Figure 4: Taughmaconnell in the Stafford Survey Map for the Barony of Athlone (1636)

Note that the masonry building that should be situated in the area labelled Dromdonal, is conspicuous by its

absence.

29

Barry, T., The Archaeology of Medieval Ireland, p. 9

15

The First Edition Six-Inch Ordnance Survey Maps were produced between 1824 and

1846. These maps are of value as they depict the landscape, with all known antiquities

at the time, prior to the reorganization of the countryside after the Famine and since

then.30

Dundonnell is depicted on Six-Inch Sheet 47 in the Roscommon series (Fig. 5).

Figure 5: Six Inch 1st edition Ordnance Survey Map of Dundonnell (1829)

1.5 – Conclusions

In short, this chapter has outlined the various sources and methods used in developing

this thesis, with the aims of the work catalogued at the outset. In essence, Dundonnell

was chosen for study in order to provide a clear understanding of a neglected

monument in the Irish landscape, and possibly to work as a blueprint for future study of

this and other neglected sites.

30

Barry, T., The Archaeology of Medieval Ireland, p. 10

16

Chapter 2-The Historical Background

2.0 – Introduction to the Chapter

Given Dundonnell Castle‟s location in the landscape of south Roscommon, it has been

deemed necessary to provide a historical treatment of the area from the period that the

first habitation could have occurred at Dundonnell during the early medieval period up

until the 17th

century, when the stronghouse on the site was probably last occupied.

This treatment has to be constructed for the sake of context, without which, the site

itself would be suspended in a historical limbo that would inform us of little. This will

be followed by a brief discussion of Taughmaconnell parish itself, and what

information we can ascribe to the immediate area for the date range in question, in

order to further extend our understanding of Dundonnell. Finally, and most importantly,

the site itself will be discussed in its historical context. Its various references in the

historical record need to be documented and analysed.

2.1 - Early Medieval Ireland (c. 500AD – c. 1100AD)

Early Medieval Ireland or Early Christian Ireland is seen in popular histories as being a

Golden Age in Irish history, halcyon days for the Irish, with the archaeological and

historiographical evidence from the period rich in tales of learned monastic centres

producing great minds and texts. Ireland was a venerable land of saints and scholars.

Men such as St Columcille and St Columbanus were instrumental in converting

northern Britain and large parts of the Continent to Christianity.31

This idealised picture

of Early Christian Ireland is one that we, as a nation (within the Republic), cling dearly

to, as an authenticator of our own identity. We have seen it as a period of great

31

For example, Cahill, T., How the Irish Saved Civilisation, (London: 1995)

17

achievement for the Irish people – something that was destroyed by the English in

1169. Such long held attitudes affected the study of archaeology in Ireland with the post

1169 period receiving little attention until relatively recently due of its association with

English domination.32

In my opinion, based on my reading, 1169 should not be viewed

as the landmark date in Irish history that some historians and popular culture have often

ascribed to it.33

There was much continuity from the early medieval period into later

times and this will be discussed in more detail below. We have to remember, however,

that the Ireland that we picture today, like all modern countries, is a recent concept.

Prehistoric and Early Christian Ireland was much less a nation, much more a swathe of

petty kingdoms and saw a constant redrawing of boundaries. Resultantly, one must

constantly refer back to a different societal makeup when making conclusions.34

The

fact that the single unit of Britain first begins to take shape with the Romans cannot be

applied to its western neighbour until at least the early modern period, and it is with this

important fact in mind that we must proceed with a history of the area. John Morrissey

highlights this situation well in his period of study, and the futility of attempting to

assign such singular entities as, in his case, the Gaelic world, or the Anglo-Irish

equivalent.35

Despite the fact that for our purposes, we must transport this analogy back

over a millennium, oddly the situation is markedly similar, in that the whole way up

32

Barry, T., The Archaeology of Medieval Ireland, pgs. 1-2; McNeill, T. E., Castles in Ireland: Feudal

Power in a Gaelic World, p. 2; O‟Conor, K., The Archaeology of Medieval Rural Settlement in Ireland,

pgs. 10-12 33

See for instance the break between Ó Corráin, D., Ireland before the Normans (The Gill History of

Ireland 2: 1972) & Dolley, M., Anglo-Norman Ireland (The Gill History of Ireland 3: 1972), along with a

similar division in the more recent New History of Ireland series, both noteworthy scholarly series, but

both divided along mainstream lines. For popular culture see The Story of Ireland (2011) RTÉ, March.

10.15pm 34

MacCotter, P., Medieval Ireland: Territorial, Economic and Political Divisions, (Dublin: 2008) 35

Morrissey, J., “Cultural geographies of the contact zone: Gaels, Galls and overlapping territories in

late Medieval Ireland”, Social and Cultural Geography, Vol. 6, No. 4, August 2005, pp. 551-566, p. 562

18

through the history of the island of Ireland until at least the early modern period, the

kingdoms are so numerous, and their stories so divergent, that to provide a single

narrative would be potentially very limited and unsatisfactory.

In the same way as the island should not be viewed as a single entity in the writing of

history for these early periods, one must also be careful not to segment the timeline too

rigidly, as to do so effectively just sends the impression that there was a period of

tranquillity and purity before, say 1169, and everything post 1169 is somehow tainted.

It‟s an unfortunate situation that even occasionally rears its ugly head in scholarly

circles also, and is something I hope this study can stay away from. Only the

banishment of such divisions in this work will provide a working framework to discuss

the history of the area and the site properly. The author is aware of the necessary

division of the site itself into phases, and as such, will attempt to keep the possible

ramifications of this division to a minimum, as regards its probable change in

ownership at least.

The modern county boundaries that divide the country into thirty two entities have not,

obviously, been in existence from the dawn of time, mostly coming into existence in

the late 16th

-century, and as such, one of the prior divisions, into various „túath‟ or

ruling dynasties, becomes very important for our purposes. The title that Orpen puts to

the period immediately before the introduction of large-scale English influence

encapsulates the attitude of the political organisation of the island at the time.36

36

Orpen, G. H., Ireland under the Normans, (Dublin: 2005), p. 1 “Anarchic Ireland: Ninth to Eleventh

Centuries”. It must be noted that this is in spite of the content of the chapter, something that will not,

however, be dealt with any further in this study.

19

The early history of Connacht, alone among the other provinces, never seemed to have

had a strong tradition of over-kingship. In this regard, instead of having one or two

large túath challenging of the superiority of the province, in Connacht we had a series

of different kingdoms and intermittent wars over personal gains, but no overarching

organisation or higher ambition than that. This can be clearly seen from the 1937 article

on the “Christian Kings of Connacht”, by Rev. Walsh, which highlights that before AD

550 only three kings of Connacht can be cited with certainty.37

If, however, we add to

this the possibility that the later annotators of annals applied the title rí Connacht to an

individual posthumously,38

it serves to show that even post-550, there may not have

been a great change in practice. It is from this slightly different attitude and

organisation that we get the situation that arose in the case of our kingdom of interest,

Uí Maine, as it was this nuanced view of the overlordship that directly led to more

power secondary kingdoms, such as the aforementioned, in Early Medieval Connacht.

The ruling family during Early Medieval times in the region of south Roscommon and

east Galway, within which Dundonnell lies, was indeed the Uí Maine, or Hy-many, to

use the anglicised form (Fig. 6).

37

Walsh, P., “Christian Kings of Connacht”, Journal of the Galway Archaeological and Historical

Society, Vol. 17, No. ¾ (1937), pp. 124-143, p.142 38

Ó Cróinín, D., “Ireland 400-800”, in Ó Cróinín, D., (ed.), The New History of Ireland 1: Prehistoric

and Early Ireland, (Oxford: 2008), pp. 182-234, pgs. 227-228

20

Figure 6: Territory of the kingdom of Uí Maine (expansion & decline taken into account)

The Uí Maine had a rich history, which stretched back to their pseudo-genealogy,

summarised by Byrne39

, and semi-mythological arrival in the south Roscommon area

under the patronage of a follower of St. Patrick, St. Grellan, in the 5th

-century AD.40

39

Byrne, F. J., Irish Kings and High-Kings, (Dublin: 2001), p. 85 40

O‟Donovan, J., The Tribes and Customs of Hy-Many, commonly called O’Kelly’s country, (Dublin:

1843), pgs. 8-9

21

Tirechán, the late 7th

-century biographer of St. Patrick, makes brief reference to the Uí

Maine and describes the founding by Patrick of a church in their territory at Fidarta or

Fuerty, Co. Roscommon.41

Kelleher provides an important point with regard to the treatment of the Uí Maine in the

annals, specifically the Annals of Clonmacnoise, in that, although the monastery was in

the territory of Mide and was patronised by the Clann Colman Móir and the Uí Fhailge,

the Uí Maine regularly receive more frequent mention than the patrons themselves.

This he argued was because “Uí Maine territory lay along the west side of the Shannon

from just south of Loch Derg to some point on Loch Ree, and on both sides of the

lower Suck, (therefore) every military force that crossed in either direction at Athlone

or that moved up or down the Shannon came through their lands or impinged upon

them.”42

In other words, the Uí Maine were one of the closest tribes to Clonmacnoise

during Early Medieval times and were able to influence affairs at the monastery as a

result.

Thus, the kingdom of the Uí Maine was a powerful one, albeit „of the second rank of

importance‟ in Connacht due to its lack of a descent from Echu Mugmedón. In spite of

this, it was probably one of the largest petty kingdoms on the island in the Early

Medieval period.43

Its location on the Shannon also made the Uí Maine powerful as it

allowed them control this vital north/south routeway. Besides this, the Uí Maine were

also very influential in other spheres.

41

Ó Cróinín, D., “Ireland 400-800”, p. 231 42

Kelleher, J. V., “Uí Maine in the annals and genealogies to 1225”, Celtica, ix (1971), pp. 61-112, p.

62 43

Walsh, P., “Connacht in the Book of Rights”, Journal of the Galway Archaeological and Historical

Society, Vol. 19, No. ½ (1940), pp. 1-15, p. 12

22

Clonfert, Aughrim (Echdruim) and most importantly, the monastic „city‟ of

Clonmacnoise, a worthy rival for the „Patrician‟ Armagh, were all situated within or

close to Uí Maine territory, with Uí Maine kings even having the privilege of being laid

to rest in Clonmacnoise.44

The prestige associated with having religious communities

on one‟s land was not lost on the Uí Maine. Rich monasteries like Clonmacnoise

provided much wealth for adjacent tribes, albeit not always at the permission of the

incumbent abbot and his monks.45

The first annalistic reference to the Uí Maine comes in 538, with the Battle of

Clóenloch, and it is with this and certain later annalistic references concerning this

group that we can glean a very interesting power that the túath possessed. Essentially,

in order for a contender to obtain the overlordship of the province of Connacht, the

kingdom of the Uí Maine had to be dominated first.46

The position of the Uí Maine in

this regard in later years was not without challenge, and by the 8th

century, the Uí

Briúin were making advances on the overlordship of Connacht, slowly wearing down

the power of the Uí Maine on their way to becoming major players on the politics of the

island. But, for a period of nearly two centuries, the Uí Maine were the kingdom that

had to be appeased in order for another túath to have safe access to the kingship of

Connacht.

44

Byrne, Irish Kings and High-Kings, pgs. 92-93 & 252 45

Kelleher, “Uí Maine in the annals and genealogies”, p. 62 46

Ó Cróinín, “Ireland 400-800”, p. 232

23

Despite the eventual demise of their useful position as a powerbroker, an inevitable

situation really, given the ebb and flow of politics in Early Medieval Ireland, the Uí

Maine were still very powerful in the region and the island as a whole in the 10th

, 11th

and 12th

-centuries. For example, it is clear that at least one member of the Uí Maine

played a prominent part in the Battle of Clontarf in 1014. In the notes section of the

Tribes and Customs of Hy-Many, O‟Donovan includes biographical information on one

Tadhg Mór O‟Kelly, son of Murchadh, and chief of the Uí Maine for thirteen years

before he fell “like a wolf-dog fighting the Danes” in the battle. Around his heroism, a

unique one as the men of the Uí Fiachrach Aidne and the Uí Maine were the only two

non-Munster based kingdoms to provide contingents in Brian Bóruma‟s army in

101447

, grew a tale that said that a fabulous animal came from the sea to protect Tadhg

Mór‟s fallen corpse from the Danes.48

This is, no doubt, a tribute to the man who was

recorded as being one of the three coimmite or attendants of Brian Bóruma at this

time.49

So while the Uí Maine no longer held the balance of power in Connacht, they

still had the distinction of having powerful friends. It is also noteworthy that the

O‟Kellys had become the dominant sept of the Uí Maine by the time of Clontarf.

The intriguing characteristic that seems to make Uí Maine territory so central to the

politics of early medieval Ireland and Connacht is just that, its geographic centrality on

the island itself. This has been previously alluded to with regard to its proximity to the

Shannon and closeness to the borders of Mide and Leinster, but is compounded even

further by the presence of a number of key land routes that passed straight through Uí

47

Byrne, Irish Kings and High-Kings, p. 243 48

O‟Donovan, The Tribes and Customs of Hy-Many…, p. 99 49

Kelleher, “Uí Maine in the annals and genealogies”, p. 84

24

Maine territory, with the most prominent being the Slighe Mhór. This meant that many

major expeditions east to west or vice versa, be it martial or mercantile in motive, had

to pass through South Roscommon.

It was Colm Ó Lochlainn who initiated the first extensive study into the Early Medieval

routeways of the island. His mapping of the routes, using mythology and folklore as his

primary sources50

, has been generally accepted as correct and built on in more modern

studies, with Michelle Comber51

and, more specifically to the needs of this thesis,

Linda Doran52

providing much needed research in the fields of communication and

transport in Early Medieval Ireland. It is clear from these studies that a very important

routeway ran across Ui Maine territory in south Roscommon during Early Medieval

times. This road entered the latter territory at a ford over the River Shannon and crossed

the Suck at Ballinasloe. Both the Suck and Shannon were navigable rivers and so it can

be said that the wealth and power of the Uí Maine was linked to their access to and

control of the route.53

Transportation and communication links in relation to the site

will be given a full discussion later in the study.

In conclusion, for the Early Medieval period, the Uí Maine (whose main grouping

became the O‟Kellys) was a wealthy and politically important tribe. Important

routeways went across their territories, within which Dundonnell was situated. Rich

50

Ó Lochlainn, C., “Roadways in Ancient Ireland”, in Ryan, J., (ed.) Féilsgribhinn Eóin Mhic Néill,

(Dublin: 1940), pp. 465-474, p. 465 51

Comber, M., “Trade and Communication in Early Historic Ireland”, The Journal of Irish Archaeology,

Vol. 10 (2001), pp. 73-92 52

Doran, L., “Medieval Communication Routes through Longford and Roscommon and Their

Associated Settlements”. 53

Ó Lochlainn, C., “Roadways in Ancient Ireland”; Comber, M., “Trade and Communication in Early

Historic Ireland”; Doran, L., “Medieval Communication Routes through Longford and Roscommon and

Their Associated Settlements”.

25

monasteries occurred along the borders of their lands. In all, the kingdom of Uí Maine

was one of the most important and largest petty territories within the province of

Connacht.

2.2 - High Medieval Ireland (c. 1100 – c. 1380)

Indeed, in the early years after 1169, the situation in Uí Maine didn‟t change all that

much, when the populist readings of the situation are placed to one side.54

The arrival

of the English initially made little impact on Connacht in general and Uí Maine in

particular. The O‟Kellys remained the kings of Uí Maine. As the 12th

-century came to a

close, the political situation in South Roscommon is ultimately preoccupied with the

wider dynastic and political struggles of the O‟Conors, and the effect these struggles

had in their individual kingdoms, than the newly colonising English. The

unprecedented stranglehold that the O‟Conors held on power in both the region and the

island stretched back to the 10th

-century.55

Therefore their influence was felt the length

and breadth of the island, something that only started to deteriorate with the ever-

increasing power of the English.

O‟Conor power would eventually be eroded, however, not just by outside forces but

also by the intricate nature of Irish dynastic politics.56

Although this was not a new

occurrence in medieval Ireland, the position of the English would have great

implications regarding the future history of Uí Maine territory, Connacht, and the island

54

Gerald of Wales, The History and Topography of Ireland, trans. O‟Meara, J. J., (Penguin, London:

1982). Giraldus Cambrensis is the first to provide a detailed account of the English arrival, however, his

propagandistic text should be treated with extreme caution. 55

Byrne, Irish Kings and High-Kings, p. 253 56

Martin, F. X., “John, Lord of Ireland, 1185-1216”, in Cosgrove, A., (ed.) A New History of Ireland II:

Medieval Ireland 1169-1534, (Oxford: 2008), p. 130

26

as a whole, although the actions that led to this occurred much more subtly than

traditional historians like Orpen may have had you believe.57

The direct dealings between the Uí Maine and the Kings of Connacht in the early years

after 1169 are mixed, with the taciturn element being central to their dealings.

The first year after the arrival of the English on the island sees the Annals of Tigernach

and Annals of the Four Masters more concerned with the retaliatory expedition of the

then high-king of Ireland, Ruaidrí O‟Conor, against the tribes of the Úa Briain and the

Dál Cais, with Uí Maine providing assistance to the high-king by raiding the lands of

Ormond, directly to their south, raiding which continued into 1171.58

Here we see the

strategic advantage that allying with the Uí Maine provided, something that could

easily be transferred to raiding expeditions across the Shannon also. If, however, we

fast forward to 1180, we see a complete reversal of fortunes for the Uí Maine, and their

locational advantages counted for little, as the O‟Conors find themselves at war with

the men of Uí Maine. In this battle we find Conchobar Maenmaige, son of Ruaidrí,

defeating and killing the king of the Uí Maine, Conchobair O‟Kelly, along with much

of his kin59

, showing how quickly relations soured.

Ruaidrí was the king of Connacht and high-king of Ireland upon the arrival of the

English in 1169, and with Henry‟s visit in 1171-72, commentators of the period, such

57

Seán O‟Faolain introduces a very interesting hypothetical model for the more gradual subjugation of

the Irish „captains‟, as he puts it, with Donal O‟Conor Sligo being negotiated into the ‟Surrender and

Regrant‟ system in the 16th century reign of Henry VIII. The earlier precursors of this system were first

to be employed in Connacht with the English „mercenary‟ involvement into interfamilial strife, such as

the turmoil that enveloped the O‟Conor in the 12th and early 13th centuries. See O‟Faolain, S., The Great

O’Neill, (Mercier Press, Cork 1997), pgs. 11-14 58

Kelleher, “Uí Maine in the annals and genealogies”, p. 102 59

Kelleher, “Uí Maine in the annals and genealogies”, pgs. 102-103

27

as Giraldus Cambrensis, asserted that this was when the English effectively claimed

dominion of the island. However, just because a number of kings offered submission

does not mean that Henry was the de facto ruler.60

In fact, if we consult the other

sources and treat Giraldus with caution, we can come to the conclusion that Ruaidrí

indeed became Henry II‟s homo, but not as a submissive tribute, rather as King of

Connacht and ruler of all but the lands that Henry and his English subjects controlled,

in return for faithful and obedient service.61

However, the nature of English private adventuring meant that ambitious men would

always push the boundaries of agreements such as the Treaty of Windsor, signed in

1175. The first violations occurred one year later, with John de Courcy setting about the

invasion of Ulster, and in 1177, when, more specifically to our needs, Miles de Cogan,

a constable of the Dublin garrison, crossed the Shannon with a force of knights and

archers to invade Connacht. It seems that this was done effectively at the behest of

Murchad O„Conor, a son of Ruaidrí‟s, „for evil towards his father‟. This act of

mercenary activity was essentially the commencement of “a process which was to

become a regular feature of English involvement in Connacht, and which was usually

very rewarding for the English in terms of both wages and booty.”62

Although the De

Cogan expedition failed, it was merely the first venture across the Shannon, often at

Athlone in Uí Maine lands, to claim Connacht, showing the concrete aims that the

English had in mind for the province.

60

Walton, H., “The English in Connacht, 1171-1333”, (Unpublished PhD Trinity College Dublin: 1980),

p. 2 61

Walton, “The English in Connacht”, p. 8 62

Walton, “The English in Connacht”, p. 13

28

The year 1178 sees the action turn to Uí Maine territory specifically for the first time.

Another act of English aggression on the part of Hugh de Lacy was attempted at

Clonmacnoise in this year. However, an O‟Conor army, presumably supplemented by

Uí Maine troops, were steadfast in their determination to repel the English, both for the

prestige Clonmacnoise held for both their kingdoms, and the possibility that capturing

Clonmacnoise may have been used as a springboard into the south east of Connacht.63

The importance that Athlone held as gateway to Connacht, along with its proximity just

inside Uí Maine lands, makes its role in the 12th

and 13th

centuries very important for

our purposes. Orpen describes Athlone as such - “Whoever held Athlone, held the key

to communications between Connaught and Meath…The Irish used to make fords by

widening the channel of a river, and even by building a sub-aqueous tochar or

causeway. Such a tochar was made at Athlone in the year 1000.”64

To complement the

ford crossing at Athlone and to concrete its position as a base for incursions into Mide,

in 1129 Tairrdelbach O‟Conor, Ruaidrí‟s father, erected a wickerwork bridge and a

caislen or castle on the eastern bank of the Shannon at Athlone. Combined, the bridge

and fortification were to prove a constant nuisance to the lands and kings of Mide, as

seen by the fact that Tairrdelbach and, later, his son Ruaidrí, constructed a total of five

bridges at the crossing in the space of twenty-nine years. This was due, in no small part

to the fact that the kings of Mide were constantly pulling said bridges down.

63

Walton, “The English in Connacht”, p. 16, citing the Annals of Tigernach 64

Orpen, G. H., “Athlone Castle: Its Early History, with Notes on some Neighbouring Castles”, The

Journal of the Royal Society of Antiquaries of Ireland, Fifth Series, Vol. 37, No. 3, [Fifth Series, Vol. 17]

(Sep. 30, 1907), pp. 257-276, p. 258

29

With the arrival of the English on the island and their interest burgeoning in Connacht,

Athlone assumed an even greater position of importance in the destiny of the province,

and by 1200, the general consensus is that the settlement of Athlone was now in the

possession of the English. The entry in the annals of Loch Cé for 1199 certainly points

towards a situation of Cathal Crobderg, the younger brother of Ruaidrí, attacking the

bódhún of Ath, which has come to be thought of as a reference to the bawn of Athlone.

At this stage there was probably a fortification on the western bank of the Shannon65

,

on the site of the fortification that survives to present day. In the long run, the operation

of the „gateway‟ of Athlone changed hands, with the control of the bridge opening up

the whole of Connacht and the north of the country to the English66

, making it a key

advantage to their aspirations of expansion into these areas.

Athlone thereafter became the site of two peace settlements on the parts of Cathal

Crobderg and the English, in 1195 and 1210. The ambitious Crobderg had aspirations

of increasing his power on the island, thus starting to raid in neighbouring eastern

kingdoms. The English became aware of his ambitions, and met him at Athlone, where

his kingship was recognised in return for him ceasing his expeditions outside of his

own borders.67

The second Peace of Athlone (1210) continued this observance, after a

few instances of instability and uncertainty in the intervening period.68

The place that

Athlone held here gives us an indication of the kind of traffic that would have passed

through Uí Maine territory, and the O‟Kellys would no doubt have been present at these

summits on their lands.

65

Orpen, “Athlone Castle: Its Early History, with Notes on some Neighbouring Castles”, p. 259 66

Martin, F. X., “John, Lord of Ireland, 1185-1216”, p. 146 67

Dudley Edwards, “Anglo-Norman Relations with Connacht”, p. 146 68

Dudley Edwards, “Anglo-Norman Relations with Connacht”, p. 151

30

However, Cathal Crobderg‟s death in 1224 resulted in twelve years of war that would

change the political geography of Roscommon beyond recognition.69

Connacht was

granted to Richard de Burgh in 1227, with five cantreds to be set-aside for the O‟Conor

king of Connacht, essentially as royal tenants. The cantreds contained basically the

borders of modern day Roscommon - Uí Maine, Tír Maine, Mag nAí, Trí Tuatha and

Mag Luirg-Uí Aillelo (Fig. 7).

The rest of Connacht was to be settled by the English.70

This action was the first real

attempt to colonise Connacht, driven partly by the belligerence of the locals but also by

the ambition of individual English lords.

The mid 13th

century O‟Conor king was Fedlimid, who was in the strange position of

being both a vassal and a king. He spent his reign trying to hold on to these vastly

reduced lands, even visiting Henry III in attempts to show himself as true to his English

masters.71

This strategy of appeasement did succeed in Fedlimid transferring his

kingship onto his son, but the diplomacy Fedlimid toiled over was to be obliterated by

said son, Áed “na Gall”.72

69

Graham, “Medieval Settlement in Roscommon”, p. 23 70

Graham, “Medieval Settlement in Roscommon”, p. 23 71

The reasons for Fedlimid‟s visits are well documented in Verstaten, F., “Both King and Vassal:

Feidlim Ua Conchobair of Connacht, 1230-65”, Journal of the Galway Archaeological and Historical

Society, Vol. 55 (2003), pp. 13-37, pgs. 20-21. 72

The nickname Áed na Gall, rather than being reference to a link with the English, refers to his

marriage to the daughter of Mac Sumarlaide, King of the Hebrides, in 1259, in the process being granted

8 score Scottish mercenaries as a dowry. These are the first recorded mention of the gallowglasses in

Irish history, thus leading to his nickname. See Lydon, J., “A land of war”, in Cosgrove, A., (ed.) A New

History of Ireland II: Medieval Ireland 1169-1534, (Oxford: 2008), p. 245

31

Figure 7: Map of the King's Cantreds

The rise to power of Áed O‟Conor in the mid 13th

century was the result of a complete

change of policy in dealing with the English.73

Áed caused havoc for the English in

Connacht and adjacent to it, during an unusual period of time when a number of

73

Lydon, J., “A land of war”, p. 247

32

ascendant, militant and nonconformist Gaelic lords began to stop the hitherto fairly

straightforward English expansion. Áed first comes up on English radars when he lined

out on the Ó Neill side against the English at the Battle of Down (1260), narrowly

coming away with his life intact. Retaliation was swift. Connacht was invaded in 1262.

Between 1262-66, however, Áed was the one in the ascendancy. He sacked Athlone

and plundered many of the English settled lands in Connacht. He succeeded his father

to the kingship in 1265, and, when the English eventually summoned him to a meeting

in Athlone in 1268, Áed merely turned up with a large host and sacked it again.74

The

now cantred of Uí Maine was again seeing much of the fighting, due to its key position

as the gateway to Connacht.

English reaction came with the construction of the royal castle of Roscommon in 1270,

with the castle plans including the postern gatehouse,75

built specifically to look out

over a royal O‟Conor crannóg on the adjacent lake. Roscommon was planned

essentially to push the O‟Conors further and further north, thus letting Connacht be

settled peacefully. The justiciar, Robert d‟Ufford, even brought troops over to help in

the defence of the construction site, such was the priority endowed on the construction

of this fortification. This was common practice in hostile border areas; understandable

given the expense required even to recruit the high standard of worker needed to build a

fortress of this kind.76

Áed turned up and soon turned on them, with the troops having

74

Annála Connacht: the Annals of Connacht, AD. 1224-1544, p. 151 75

Murphy, M. & O‟Conor, K., Roscommon Castle: A Visitor’s Guide, (Roscommon County Council:

2008), p. 10 76

Warner, P., The Medieval Castle: Life in a Fortress in Peace and War, (Penguin Classic History:

2001), p. 151

33

to retreat back across the Shannon near Carrick-on-Shannon.77

The new castle was

knocked down, and from then on a near constant raiding occurred on towns and castles

across Roscommon, as Áed pushed the English out of east Connacht and the King‟s

Cantreds. This year saw the destruction of a Crown castle at Onagh.78

In 1272, he burnt Athlone again and destroyed the bridge, along with Roscommon and

Rindown and raided even as far as Granard. This belligerence was to come to an abrupt

end, however, on 3rd

May 1274, when Áed died, plunging Connacht into civil war and

undoing much, if not all, of his previous work in removing settlers from the lands west

of the Shannon.

Despite this being a physical end to the troublesome reign of Áed O‟Conor, the

financial ramifications recorded for the subsequent repairs and maintenance of the three

traditionally cited royal castles, along with the continued payment of rent on a number

of damaged sites79

show two things. First they show the extent of the headache and

worse that Áed provided for what had been, prior to this, a fairly uncomplicated

settlement of Connacht. But it also shows how determined the English were in holding

on to their lands and castles.

With the turn of the century, the once sustained push to hold Connacht was starting to

fade, with the near incessant fighting and destruction leaving the royal lands in Athlone

77

Ann. Conn., pp. 155-157 78

The Annals of Clonmacnoise: Being annals of Ireland from the earliest period to A.D. 1408, p. 249 79

Calendar of documents relating to Ireland, 1252-1284, (vol. 2), Sweetman, H. S. (ed.), (Nendeln,

1974), p. 550, entry 2329, year 1284: “Moylonach was wont to answer for 6l. 5s. as is contained

therein…”, this is in spite, obviously, of the fact, that it had been destroyed 14 years previous and with no

records of a bill for its repair with us, it has to be assumed that it was still in a state of disuse or reduced

function, but was still deemed important enough to be paying rent.

34

and Roscommon worth nothing, and with conflicts to fight elsewhere, resources to man

Connacht and the other hostile kingships were simply not there anymore.80

The three principles that, according to J. A. Watt, comprised the policy that the English

had in Ireland, had failed.81

Ireland went from being a source of revenue for the Crown

to being a charge on it,82

and the uncertain nature of the legal system83

being two of the

more prominent negatives that could be posited about the English continued dominion

over the island, pushed the situation to the brink of collapse.

It was during this period that many regions, such as Uí Maine, returned to their

previous owners, so that by 1347, by Nicholls calculation, Connacht was effectively

lost back to the Irish.84

The subsequent growing power of the O‟Kellys of Uí Maine can

be dated to around this time, with their influence in the wider political field lasting until

the early 16th

century.85

2.3 - Late Medieval Ireland (c. 1380 – c. 1650)

With the steadily decreasing influence that the English were having on the island, and

in our case eastern Connacht, the vacuum their absence created effectively allowed for

80

Lydon, “A land of war”, pgs. 250-251 81

Firstly, the colony must be self-sufficient and also provide resources for crown needs in the form of

revenue, men and provisions. Secondly, the legal system, the structures of same, and the institutions of

both church and state were to be the same in both the parent kingdom and the colony, thus helping with

law and its enforcement, along with general compliance and conformity. The third principle is an

offshoot of the second and concerns the supervision of the colony, which manifested itself in the real

world with the Dublin administration, who could act directly on behalf of the King himself, Watt, J. A.,

“The Anglo-Irish colony under strain, 1327-99”, in Cosgrove, A., (ed.) A New History of Ireland II:

Medieval Ireland 1169-1534, (Oxford: 2008), pgs. 374-375 82

Watt, “The Anglo-Irish colony under strain, 1327-1399”, p. 376 83

Watt, “The Anglo-Irish colony under strain, 1327-1399”, p. 377 84

Nicholls, K., Gaelic and Gaelicised Ireland in the Middle Ages, (The Gill History of Ireland 4: 1972),

p. 18 85

Nicholls, Gaelic and Gaelicised Ireland, p. 148-150

35

a resumption of the age old hostilities that gripped the politics of Connacht. The newly

ascendant O‟Kellys of Uí Maine capitalised on this situation in the 14th

century and

their strategic siding with the O‟Conor Roe.86

Indicative of this newly elevated status

that the O‟Kellys now possessed is seen with the production of the Leabhar Uí Maine

in 1394, right in the middle of wider dynastic struggles between two branches of the

O„Conors, showing it to be a profitable period for those who were allied to one branch

or the other.

The internal struggles of the O‟Conor was to preoccupy Connacht until the very end of

the 15th

century, when even this battle was growing stale and the two branches seemed

more content with putting their efforts into fighting one another than succeeding as a

political force.87

Two branches of the Burkes, those of Clanricard in the south, and the

MacWilliam Burkes in Mayo and Sligo, began to take temporary control in their

respective regions, as new alliances were formed, and older ones failed. Again, as is the

wont of the cyclical politics of túath based Ireland, the O‟Kellys in this period were

again on the decline, due mainly to as succession of weak central chiefs, undermining

the power they held in their own territory and the respect they commanded outside.88

Native politics was to get another overhaul, however, with the arrival of Henry VIII on

the English throne.

The change of policy that Henry VIII‟s Tudor government enacted in Ireland was, in

essence, the beginning of a struggle to return Ireland to the position that his

86

Cosgrove, A., “Ireland beyond the Pale, 1399-1460”, in Cosgrove, A., (ed.) A New History of Ireland

II: Medieval Ireland 1169-1534, (Oxford: 2008), p. 577 87

Quinn, D. B., “ „Irish‟ Ireland and „English‟ Ireland”, in Cosgrove, A., (ed.) A New History of Ireland

II: Medieval Ireland 1169-1534, (Oxford: 2008), p. 625 88

Nicholls, Gaelic and Gaelicised Ireland, p. 150

36

predecessors had held in Ireland. However, the real period of re-colonisation or

plantation of the country occurred in the reign of Queen Elizabeth I. The first

Elizabethan grant of land in Roscommon occurred in 1572 when Sir Edward Fitton,

appointed first president of Connacht in 1569, wrote to Lord Burghley for fifty men to

put the land of Roscommon to good use.89

Although, this was the first effective land

grant for the region in a number of centuries, it was simply perpetuating a policy of

colonisation that had been English practice in all their colonies, from their conquering

of the Welsh onwards. It was a theory of government that relied upon pinning down a

country with castles whose owners had a personal stake in the success of the venture.90

Fitton was merely one more wilful adherent to the policy.

Due to various problems from the locals91

, the situation that Fitton had gotten himself

into was one that he was adjudged not capable of dealing with one his own. Resultantly,

in 1576, Nicholas Malby was sent with a force by the queen in order to reduce the Irish

to submission. To these ends the lord deputy, Sir Henry Sidney, aided him and over the

winter of 1576-77, the duo succeeded in their objectives. After some minor incidents,

Malby had secured both the castles of Athlone and Roscommon, and the abbeys of

Roscommon and Boyle, so that in 1577 he had a list written up of the rents now due to

the queen in Connacht.92

He furthered his objectives in 1578 by requesting off the

queen that he be allowed to wall the settlements in Athlone and Roscommon and

garrison the latter. With all of this requested and granted, Malby, by now governor of

89

Calendar of State Papers relating to Ireland of the reigns of Henry VIII, Edward VI, and Elizabeth,

1509-[1603]. Hamilton, H. C. (ed.). 11 vols. P.R.O. (London: 1920-31), 1509-73, p. 424 90

Warner, P., The Medieval Castle: Life in a Fortress in Peace and War, pg. 137-138 91

Cronin, T., “The Elizabethan colony in Co. Roscommon”, in Murtagh, H. (ed.), Irish midlands

studies-essays in commemoration of N. W. English, 107-20, (Athlone, Old Athlone Society: 1980), p.

107-108 92

Calendar of State Papers relating to Ireland, 1574-85, p. 126

37

Connacht, was in a position to settle the province, populated, by on large, by the

horsemen who were under his charge when he commanded in Ulster.93

Malby may have portrayed the colonisation of Connacht in fairly rosy terms but the

reality was slightly more difficult. With his death came the appointment of Sir Richard

Bingham as governor of Connacht. Hostilities were rife in Roscommon as Bingham

took over, with the mere travel from Athlone to Roscommon a hazardous undertaking.

However, the problem lay much less with the fortifications themselves than with the

routeways. Bingham‟s men held a substantial number of strongholds in the region,

Athlone, Curraghboy, Lisdalon, Roscommon, abbeys Tulsk and Boyle being the most

important, but moving between these was a perilous maze in 1596, when

Castlesampson, Cornegee, Coolegarry and Moyvannon, all in the boggy and already

naturally treacherous hinterland of Athlone, Uí Maine lands, were in Irish hands. That‟s

even before moving northwards, closer to Roscommon,94

where the situation became

even bleaker.

This, however, was merely a sign of things to come. Roscommon Castle was besieged

in February and March 1596. The town destroyed, discontent spread throughout the

English garrisons of the county and mutiny seemed to be a very real possibility, in the

case of the soldiers of Athlone at least.95

Connacht was already being raided incessantly

by the O‟Donnells of Tír Connell as early as 1595, and the civil unrest only brought

more. A count of the buildings captured by the Irish in Roscommon was compiled in

93

Cronin, “The Elizabethan colony in Co. Roscommon”, p. 111 94

Cronin, “The Elizabethan colony in Co. Roscommon”, p. 116 95

Cronin, “The Elizabethan colony in Co. Roscommon”, p. 117

38

May 1596, and it reached, at its height, a total of nine abbeys, forty-four castles,

eighteen strongholds and nineteen forts of earth. In spite of this, all the major

strongholds were still in English hands.96

However, as before, this made

communication nigh on impossible, as the Nine Years War started to gain pace.

The results of the Nine Years War are obviously outside the scope of this history, save

to say that in the aftermath, Roscommon was never officially planted by English

settlers, adjudging the difficulty that they encountered in the years after 1572 to be

reason enough not to attempt it again. The Gaelic families of the county retained

control over the majority of their lands until their confiscations in the mid-to-late 17th

-

century97

, with English and old-English settlers at a premium. The likes of the

O‟Conor, MacDermot, and O‟Hanly in the north, and the O‟Kelly and Mac Keogh in

the south of the county were still the principle landowners at the end of our period.98

The ownership of the lands in South Roscommon, specifically Taughmaconnell, were

roundly transferred on the most part to Robert and Lady Talbot in the Cromwellian

Grants of 1654-1658, but in a number of instances, Mac Keoghs held on to some of the

lands of their kin99

, the Mac Keoghs being a sept of the Uí Maine that split from the

O‟Kellys in the 15th

century100

, in a practice probably akin to what Nicholls describes

occurred at May Day celebrations up until at least the mid 17th

-century.101

96

Cronin interestingly adds to his list of strongholds still in English hands in 1596, „probably

Dundonnell‟, but he doesn‟t provide a reason for his mention. See Cronin, “The Elizabethan colony in

Co. Roscommon”, p. 118 97

O‟Conor, K., “English settlement and change in Roscommon during the late sixteenth and seventeenth

centuries”, pp. 189-203 98

O‟Conor Don, C.O., The O’Conors of Connaught, (Hodges & Figgis, Dublin: 1891), pgs.213-243 99

Cromwellian Grantees 1654-1658: County Roscommon Tagh Mc : Connell Parish Athlone Bar:,

recorded in Taughmaconnell: a history, (Athlone: 2000), pgs. 19-26 100

Moore, M. (ed.), Taughmaconnell: a history, (Athlone: 2000), p. 16 & pgs. 246-248 101

Nicholls, Gaelic and Gaelicised Ireland, pgs. 61-62

39

2.4 - Taughmaconnell parish

As the modern parish of Taughmaconnell is situated in what was historically Uí Maine

territory, it is important to discuss its history throughout this period also, to understand

Dundonnell‟s history locally. According to a general history of the parish, the earliest

mentions of this area in the annals would not have used Teach Mac Conail, rather

utilising Magh Finn, anglicised to Moyfinn. The first recorded use of Teach Mac

Conail only occurs really with the publishing of the Stafford Survey in 1636.102

Interestingly, one of the first recorded mentions of the former, Magh Finn, occurs in

Tírechan‟s Life of Patrick, and a story concerning the confusion between the graves of a

pagan and a baptised Christian that Patrick put right.103

Parts of Moyfinn were parcelled off to the monastery of Clonmacnoise in donation by

the Uí Maine at an, unfortunately unknown, early date, showing again the close links

between the Uí Maine kings and the monastery. Given the area‟s relatively close

proximity to Clonmacnoise bolsters this theory, and there is a relatively widely held

local belief that Moyfinn was always „church lands‟.

With the demise of the power of Clonmacnoise, we get a division and distribution of

the lands to the various diocesan powers that grew up in the place of dying

monasticism. This transfer can be dated with some accuracy to a meeting of the local

bishops in Tuam in 1210, with the area being divided up between Clonfert, Tuam, and

Clonmacnoise diocese, Clontuskert Priory, Aughrim Priory and the Franciscans of

102

Taughmaconnell: a history, p. 36 103

Taughmaconnell: a history, p. 35

40

Athlone.104

Notable religious communities that were active in the area included the

monastic cell that was present as early in the 9th

century in the townland of

Bellaneeny,105

approximately 1 kilometre due north of the site of Dundonnell Castle,

with the 13th

century Augustinian cell at Cloonoghill, or Cloonlawhill106

lying within 4

kilometres east of same. Combined they show some signs of the kind of associated

buildings that regularly go hand in hand with castle location on the British Isles,

however, more on that later.

The later Middle Ages for Moyfinn can be summarised with noting the further

transferral of the lands, and the opposition they faced from the previous occupants.

Connacht was granted to Richard de Burgh in 1227,107

with the King‟s Cantreds being

the only lands left aside for the O‟Conor king. 1250 saw a whole succession of land

grants, reinforced by the policy of private castle building that led to the construction of

Roscommon castle and Onagh among others.

By the 14th

century, however, the tide was turning, and the resurgent Connacht tribes

were making English settlement unviable, only for the new wave, under Malby, and

later Bingham, to start the process all over again. Croinin asserts that two of Malby‟s

Ulster cavalry, John Willing and John More went on to hold parts of

104

Taughmaconnell: a history, p. 18 105

Egan, P. K., “The Carmelite Cell of Bealaneny”, Journal of the Galway Archaeology and Historical

Society, Vol. 26, No. ½ (1954/1955), pp. 19-25; Archaeological Survey of Ireland, 2010, National

Monuments Service [online], available: http://webgis.archaeology.ie/NationalMonuments/flexviewer/

[accessed 9th

June 2011], Ref. No. RO051-025001 106

Moore, Dr. A., “The Cell of the Canonesses Regular of St. Augustine at Clonoghill”, Old Athlone

Society (1969), pp. 15-16; Archaeological Survey of Ireland, 2010, National Monuments Service, Ref.

No.‟s RO051-061001, RO051-061005, RO051-061007 107

Which only became politically effective in 1235

41

Taughmaconnell,108

while the local history comes to the conclusion that the Dean of

Clonmacnoise leased the townland of Kilkenny in Taughmaconnell to one Brabazon109

,

presumably the same Anthony Brabazon, who led Malby‟s Ulster horsemen. The

hostility to the colonists was also well recorded for the area, with both Castlesampson

and Brideswell referred to in the State Papers for May 1596 as being held by the

Irish.110

The 1636 Stafford Survey provides us with the first detailed map of the

parish111

, while the Cromwellian Land Grants shows us the extent of the land

confiscation in Taughmaconnell between 1654-1658, while also revealing the names of

the landowners that were dispossessed. High on the list of dispossessed families for

Taughmaconnell was the Mac Keogh sept of the Uí Maine.112

2.5 - Historical references to Onagh

To conclude this chapter, it is important to record the historical references made to the

castle of Onagh, in order to analyse later if Dundonnell could represent it on the

physical plane. The earliest reference made to Onagh occurs in the Annals of

Clonmacnoise and the Annála Connacht for the year 1236, the year of it‟s founding.113

108

Cronin, “The Elizabethan colony in Co. Roscommon”, p. 115 109

Taughmaconnell: a history, p. 13 110

Cronin, “The Elizabethan colony in Co. Roscommon”, p. 116, citing P.R.O. London, S.P. 63/189,

X/K4048 111

Stafford Survey (1636), Fig. 1, in Taughmaconnell: a history, p. 14 112

Who Dr. Moore believes to have been the possible owners and habitants of Dundonnell Castle in the

early 17th

-century. See Moore, Dr. A., “Mac Keoghs of Moyfin”, [Continued], Old Athlone Society

Journal, Vol. II, No. 5, (1978), pp. 56-70, p. 60 113

“Phelym o‟Connor with an army came to Connaught again and marched on untill he came to John‟s

house, tooke all the spoiles of the town and Llands thereof, and left nothing that they could take or see,

from the doore of the castle forth. ffelym‟s camp lay at the market cross of the towne, many of the

meaner sort of ffelym‟s army were Drownded in the pudle of that towne, he left much of the small cattle

of the said prey…The castle of Ullemme Wanagh was founded.”, Annals of Clonmacnoise, p. 235; “The

castle at Onagh was built at this time as a stronghold against the men of Connacht”.Annála Connacht, p.

60

42

This is followed up by an entry in the English Calendar of Documents relating to

Ireland for Nov. 16, 1245, showing it to have been transferred from Adam Buse into the

stewardship of John Fitz Geoffrey, thus making it a possession of the Crown, and

elevating its importance along with the castles of Rindown and Athlone.114

We must return to the Annals of Clonmacnoise for the next mention of Onagh castle,

this time concerning its destruction at the hands of Áed O‟Conor in 1270115

along with

Rindown, in an attempt to push all vestiges of English settlement beyond the borders

dictated by the boundary of the Shannon. However, indicative of the English

determination to retaining a foothold in Connacht comes with an entry from 1284,

which shows how, despite the expense needed to repair and maintain the damage

wreaked by Áed up until a decade previous, these fortifications were not abandoned

outright.116

This shows that Onagh was still paying rent in 1284, in spite of the lack of

any written evidence for it having been restored. Added to this is the fact that, in the

dealing of the rent giving of the various fortifications and castles for Connacht in

1284, the only three castles mentioned were that of Athlone, Rindown and Onagh,

highlights just how little control the English now possessed in the troublesome King‟s

Cantreds and Connacht, something that was only going to shrink to nothing over the

next century.

114

“2792. Mandate to Philip de Interberg, constable of the K.‟s castle of Limerick and of Castle Coning,

to deliver those castles with the counties of Munster and Limerick, to John Fitz Geoffrey, justiciary of

Ireland,…John de Cravill, of the castle of Athlone, with the bailiwick thereto belonging…Robert de

Capella, the castle of Randon…Adam Buse, the castle of Mayllonach.”, Calendar of documents relating

to Ireland, p. 417, entry 2792, year 1245 115

“o‟Connor immediately tooke and brake down to the earth the Castles of Athengaille, the Castle of

Sliew-Louth, and the Castle of Kilcolman, alsoe he burnt Roscomon, Rwyn-Dwyne als Teadoyn and

Vllemanagh…”,Annals of Clonmacnoise, p. 249 116

“Moylonach was wont to answer for 6l. 5s. as is contained therein ;”,Calendar of documents

relating to Ireland, p. 550, entry 2329, year 1284

43

Chapter 3 – Physical description of the site

3.0 – Introduction

The aim of this chapter is to describe the site at Dundonnell in detail. This description

is aided by a detailed topographical survey of the site, plans of different floors of the

stronghouse and accompanying photographs. As noted above, a detailed analysis of

Dundonnell has never been carried out (see 1.3). In particular, O‟Conor has noted that

detailed plans of castles of all periods are lacking in Ireland and are badly needed to

bring their study onto a more scientific basis.117

The production of a top-rate plan and

description of Dundonnell is one of the aims of this thesis (see 1.2).

In short, the site consists of a bivallate oval enclosure within which can be seen a

three-storey stronghouse.

3.1 – Siting

The site girdles a very low knoll in undulating pastureland. The First Edition Ordnance

Survey Six-Inch Map (Sheet 47 in the Roscommon Series) indicates that the latter

knoll was really a promontory of dry land jutting out into bogland (Fig. 5). This

bogland lay to the east, south and west of the site originally and was reclaimed in the

late 19th

and early 20th

centuries.

117

O‟Conor, K., “Castle Studies in Ireland”, pgs. 329-31

44

Figure 8: Dundonnell 1829 Ordnance Survey Map (Figure 5 repeated)

Natural rock outcrops can be seen in the north-western quadrant of the site. The

interior of the site is mostly under pasture today. However, the banks and ditch of the

earthwork are covered with gorse, blackthorn, brambles and hawthorn. This meant that

these bushes needed to be cleared by slash-hook and chainsaw before planning with

the Total Station could take place. In all, this labour preparing the site for planning

took approximately eighty hours to complete.

3.2 – Description of the Site (Fig. 8)

3.2a – The Earthwork

The earthwork consists of an oval enclosure, which is defined along its edges by two

earthen banks with an intervening ditch. This enclosure has internal measurements of

48m-49m north / south and 40m-41m east / west. The interior of the site is raised 1m-

2m in height above external ground level.

The inner bank has an overall width of 7m-7.5m. Its top is 0.9m-1.2m in height above

the interior of the site today, while it is it is 3.2m-3.7m in height above the base of the

45

ditch. Recent gaps can be seen cutting through this bank on its south-western and east

south-eastern sides. However, the 4m wide gap on the north-eastern side of this bank

appears to be the location of the original entrance, as it is linked to a causeway across

the fosse and thus not at the same level as the bottom of the ditch.

The ditch between the inner bank and outer bank is a full 9.5m in width, while its base

is 0.8m-1.2m in depth below external ground level. This ditch can be seen girdling the

site on its southern, western, northern and east north-eastern sides. However, the ditch

has been filled in on the eastern to south-eastern sides of the site and today a laneway

runs across these sides.118

The outer bank skirts the ditch on the southern, western, northern and north-eastern

sides of the site along the edge of the ditch. This bank is 5.5m-6.8m in overall width.

Its top is 1.75m-2.1m above the base of the fosse and 0.7m-1m in height above

external ground level.

118

Archaeological Survey of Ireland, 2010, National Monuments Service, Ref. No.‟s RO051-025010,

RO051-102----

46

Figure 9: Total Station of the earthwork,with the stronghouse situated in the north-western quadrant.

(Note the incomplete nature of the southern and eastern sections of the outer bank are incomplete due to the

heavily overgrown nature of the site at large.)

47

Figure 10: Reconstruction of what may have existed atop the earthwork at Dundonnell c. mid 13th-century,

courtesy of Carrie O'Malley (2011)

Note the narrow causeway [4m wide] may have meant that the main defensive and administrative building

may have been a timber tower on the summit. There may also have been a number of buildings on the summit

for domestic and administrative purposes also.

3.2b – The stronghouse

A defended, originally three-storey masonry building can be seen within the north-

western quadrant of the interior of the site. As noted, this building has been variously

described as a „house‟, „stronghouse‟ or „fortified house‟ (see 1.3).

It will be argued below that this structure is best described as a stronghouse (see 7.2).

At present, however, to facilitate the argument in later chapters, a physical description

of this building will now be given.

48

The building is rectangular in shape, aligned north-west / south-east on its long access

and is constructed of mortared limestone. It has three storeys - a ground floor, first

floor, and an attic. The building has internal measurements of 12.45m north-west /

south-east by 6.9m south-west / north-east. The average width of the structure‟s walls

are 1.3m – which is surprisingly wide. The house had at least six rooms, based on the

number of fireplaces but probably more. The lack of internal masonry partitions

suggests that these were of timber originally. Evidence suggests that both the interior

and exterior of the building was originally plastered. Diagonal-shaped chimneystacks

can be seen on the apex of each of the two gable walls of the building. It is locally

believed that quantities of the building‟s stone were removed in order to build the

boundary wall of a cemetery in Taughmaconnell at the turn of the 20th

century. If such

materials were indiscriminately removed at such a recent date, it may be able to

account for some of the materials that don‟t survive on the site such as the lintels,

dressed stone, and the roofing material.

The Ground floor (Fig. 9)

A splayed gun-loop occurs in each of the four corners of the building at this level. One

broken-out, originally very narrow, window (no more than 0.2m) can be seen on the

north-eastern wall of the building. One splay of another window occurs to its south-

west. Much of this north-eastern wall is gone. It is presumed that the original

entranceway to the stronghouse occurred here, as it is the only logical location for it.

49

Plate 2: North-eastern wall, largely broken out

(Also suspected location of entrance)

The long south-western wall of the building also has the remains of two originally

narrow (again no more than 0.2m in original width) broken-out windows along its

length. The south-east gable wall has the remains of a large fireplace, where the lintel

has been removed. This is the largest fireplace in the house (being 2.8m wide). This

fact and its ground-floor position suggests that the kitchen of the structure were

situated here. The splays of a broken-out, narrow window can also be seen on the

north-western gable wall at this level.

50

Plate 3: Ground floor fireplace and kitchen location on south-eastern gable wall

(Note also the first floor fireplace to the left of the main flue, presumably serving a bedroom on the first floor)

Plate 4: Gunloop on south-western wall (interior

view)

Plate 5: Gunloop on south-western wall (exterior

wall)

51

In summary, the gunloops and very narrow windows give a defensive appearance to

the ground floor. The doorway also occurred at this level. It will be argued below that

this entrance was covered by a machicolation at roof/battlement level. Also, the

existence of the largest fireplace in the house at ground floor level suggests that the

kitchen of the building occurred at the ground floor‟s south-eastern end. Its location at

ground floor could be due to fire-safety reasons. Furthermore, the size of the ground

floor would suggest that it was sub-divided. The fact that no fireplace can be seen at

the north-western gable of the ground floor may be an indication that this part of the

building was used for non-domestic purposes, such as storage, a common occurrence.

For example, the ground floors of late medieval tower houses (some of which were

contemporary with the stronghouse at Dundonnell) were mostly used for storage.119

Again, there was presumably a wooden stairs originally leading up from the ground

floor to the first floor. This may have occurred in the kitchen area. Alternatively, these

stairs could have been located in a small entranceway lobby area immediately inside

the doorway. This area would have separated the kitchen from the storage area,

meaning that the ground floor of the stronghouse was originally divided into three,

rather than two, rooms.

119

Leask, H. G., Irish Castles and Castellated Houses, (Dundalk: 1941), p. 79

52

Figure 11: Ground Floor plan of Stronghouse

The First floor

The joist holes and corbel to support the timber floor of this first floor can be seen in

the walls of the house.

Plate 6: Joist holes and corbel on south-western wall interior

53

The long south-western wall of the first floor has the remains of three rectangular

broken-out windows (each about 1m in width originally) along its length. Each of

these windows has an alcove behind them. Presumably these alcoves originally had

wooden seats within them, allowing people to sit by these lights.

Plate 7: Windows on south-western wall exterior

(Centre window now obscured by ivy)

As noted, much of the long north-eastern wall of the stronghouse has gone and been

pulled down. Nevertheless, a similar window to the above can be seen at the north-

western end of this wall. Furthermore, the left hand splay and right-hand splay of two

other windows survive at this level along this wall. It must be added that there is

evidence for a gunloop in the splay of the south-easternmost window on the length of

wall. This suggests that the first floor of the stronghouse was lit by six rectangular, 1m

wide, windows, meaning that this level was far better illuminated than the ground

floor.

54

A large fireplace occurs at this level in the north-western gable wall. This fireplace is

about 2m in width and is almost as large as the kitchen fireplace. Splayed gunloops are

positioned on either side of this fireplace feature. A smaller fireplace can be seen on

the north-eastern end of the south-eastern gable at this first-floor level. Another

splayed gunloop can be seen on the opposite, south-eastern end of this same wall.

Plate 8: North-western gable wall interior

(Gunloops positioned on either side. Note also attic floor fireplace in top left of the photograph)

55

In all, while there is evidence for four gunloops at this level, this first floor is far less

defensive than the ground floor. Here domestic comfort takes precedence over

defence, rather like early 13th

-century hall houses and hall keeps.120

The builders of the house decided that due to its height above ground level, there was

less need to defend this first floor. The very large fireplace in the north-western gable

wall and the large rectangular windows suggest that its main living room and hall of

the building lay at this end. The existence for a smaller fireplace on the south-eastern

gable suggests a bedroom area, possibly the main bedroom in the house as it was

easier to access than the attic bedrooms. Therefore, it is argued that a hall-like area and

a goodly-sized bedroom existed at this level. Presumably a corridor with a stairs in it,

giving access to the ground and attic floors separated these two rooms.

Figure 12: First floor plan of the stronghouse

120

Sweetman, D., The Medieval Castles of Ireland, p. 89, see, for example, the hall house at Tomdeely,

Co. Limerick

56

The attic floor

The south-eastern gable wall of the attic has two splayed gunloops visible within it.

The opposite north-western gable wall has a fireplace located towards its south-

western corner end and a splayed gunloop exists on its north-eastern end. No windows

occur in the gable walls. It must be presumed that, now disappeared, dormer windows

in the roof of the building lighted this level, as no other option exists.

Presumably this level was used for sleeping quarters and it was more than likely

divided into at least two bedrooms. It is uncertain as to what material was used to roof

the house. It could have been straw or reed thatch but the defensive nature of both the

building and the site suggests that this would not have been the case as raiders could

easily have set it on fire. This leaves us with the possibility of either slate, stone slates

(used as slabs) or wooden tiles being used to roof the house.121

There also appears to have been a machicolation along the eastern corner of the house

at roof level. This seems to have covered this eastern corner but also, on analogy with

contemporary tower houses and fortified houses, the postulated doorway at the south-

eastern end of the north-eastern wall.122

A doorway led from the attic space out onto

the machicolation. It is possible that there was also a wall walk and battlement along

the long walls of the building but there is no evidence for this surviving and, so, it

could equally be said that they did not exist on the site. It is also possible (if not

121

Leask, Irish Castles and Castellated Houses, p. 87 122

Leask, Irish Castles and Castellated Houses, pgs. 19, 98, & Fig. 51, North Elevation of Clara

Castle, p. 84

57

probable) that an opening in the wall allowed the machicolation to be used by

defenders standing in the attic space itself.

Plate 9: Evidence of machicolation on south-eastern gable wall

(Which would originally have been extant over to the north-eastern wall, covering the entrance in case of

attack)

Plate 10: Diamond-shaped chimneystacks

(as discussed in 3.2b)

3.3 – Conclusions

58

Therefore, this description shows that the site consists of an enclosure with a defended

stronghouse within it. Some attempt was made to understand how the rooms within the

stronghouse functioned when it was in use. O‟Conor has argued that castles of all

dates in Ireland were surrounded by timber and cob agricultural and administrative

buildings, as these places functioned as fortified estate centres.123

It is probable that

future excavation or geophysical survey at the site will find evidence for such

buildings adjacent to the stronghouse within the enclosure.

Figure 13: Reconstruction of the south facing walls of the stronghouse at Dondonnell, courtesy of Carrie

O'Malley (2011)

123

O‟Conor, K., The Archaeology of Medieval Rural Settlement in Ireland, pgs. 26-35

59

Figure 14: Reconstruction of the north walls of the stronghouse at Dundonnell, courtest of Carrie O'Malley

(2011)

60

Chapter 4 - The Siting Chapter

4.0 - Introduction

The location of Dundonnell in the surrounding landscape has been muted in the earlier

chapters, but this chapter will show just how the environment interacted with the site,

along with suggesting some of the decisions that may have resulted in Dundonnell

being placed where it was. This will be discussed chronologically from the first

possible phase, the postulated ringfort, followed by the postulated 13th

-century

ringwork castle and finally the masonry building that exists on its summit, dating to

the 16th

-or 17th

-century.

4.1 - Ringfort Siting

If we begin with the postulated ringfort124

, something that the archaeology to date has

deemed probable, we are next to encounter a contradiction between the two scholars to

have dealt with the castle in most detail to date, Graham and O‟Keeffe. Placename

evidence, used by both scholars, could influence the debate for or against the first

phase being a ringfort. Graham cited Orpen, who showed that the likes of dún or liss

could be added to a townland name after the erection of Norman fortifications125

,

while conversely, O‟Keeffe used the dún element of a place name to show its Irish,

pre-Norman origins,126

but not specifically with regard to Dundonnell itself. For my

124

Both Graham, in his 1988 work “Medieval Settlement in County Roscommon”, p. 28, and

O‟Keeffe, in his 1998 work “The Fortifications of Western Ireland, AD 1100-1300, and Their

Interpretation”, p. 190, believe Dundonnell may have had a ringfort as its initial phase, however, their

later conclusions on the site‟s development are, thereafter, completely polarized. 125

Orpen, G. H., “The mote of Lisardowlan, County Longford”, The Journal of the Royal Society of

Antiquaries of Ireland, Fifth Series, Vol. 40, No. 3, [Fifth Series, Vol. 20], (Sep. 30, 1910), pp. 223-225,

p. 225 126

O‟Keeffe, “The Fortifications of Western Ireland, AD 1100-1300, and Their Interpretation”, p. 188

61

part, I will attempt to provide the evidence that may signal it to be a ringfort also,

something that can only be answered conclusively through excavation.

Firstly, the site‟s location in relation to other ringforts in the area is interesting.

According to the Ordnance Survey Discovery Series 2nd

Edition, there are six ringforts

present in a 9km² area around the site, with the closest one located approximately 1km

north of Dundonnell.127

Stout highlights that the ringfort density in the area of South-

east Connacht is adjudged to be high in relation to the rest of the country.128

In this

alone, the likelihood of Dundonnell sitting on a ringfort is not beyond possibility.

Indeed there are areas of even higher density of ringfort distribution in the townlands

adjacent to Dundonnell townland. This density of distribution was inevitably linked to

defence and communication, as well as the more prosaic functions that would have

been ascribed to the smaller ringforts as little more than cattle enclosures. The

importance of, what Stout calls, overlapping „visual territories‟ is paramount to

defence in early medieval Ireland129

, and the site at Dundonnell has that in abundance.

The site itself is situated at approximately 54 metres above sea level, with the majority

of surrounding ringforts visible on the high ground to the south and east.

4.2 - Ringwork/Stronghouse Siting

If we entertain the possibility that a 13th

-century castle was built on the site of

Dundonnell, what aspects of the surrounding landscape would have become important

when choosing its location? What aspects of the site would make it a location to put a

127

Ordnance Survey Discovery Series, 2nd Edition, 47, with Dundonnell at Grid Ref. M 895 381;

Archaeological Survey of Ireland, 2010, National Monuments Service, Ref. No.‟s RO051-046----,

RO051-051001, RO051-050001, RO051-070----, RO050-024----, RO051-024. 128

Stout, M., The Irish Ringfort, (Four Courts Press: Dublin: 1997), p. 81-82 & Fig. 14 129

Stout, The Irish Ringfort, p. 20-21, especially Fig. 1

62

fortification to control and administer a region that in this period was very hostile?

Firstly, we must analysis the definite features of the landscape that could have been

favourable to constructing a castle there. In this regard, the features that would be

favourable for the construction of a ringwork should also be viewed with regard to the

stronghouse phase, as the factors that could have led to Dundonnell being chosen in

the 13th

-century are not altogether different from the masonry phase. The premise

behind the two phases is identical, in the sense that both would have been constructed

as private, fortified residences, positioned to control the surrounding area, during

hostile periods of time in South Roscommon.

There are two water sources in very close proximity to the site, and although the

importance of a water supply is obviously essential, Creighton leaves us in no doubt

about location choice in relation to water, with reference made to a statistical analysis

of 423 English castles and their water sources.130

In this regard, therefore, Dundonnell

is very well serviced with what is a very basic need to any permanent settlement, and

it would have served the postulated ringfort also.

The next feature that would have been favourable to the choice of Dundonnell as the

site of the 13th

-century ringwork is its position in relation to existing communication

routes. I am, of course, referring to the medieval road network that served Connacht in

the medieval period. The first routeway, and the one most directly associated to

Dundonnell, is „Route 9‟ from Linda Doran‟s study of medieval routeways through

Longford and Roscommon. She describes „Route 9‟ as being a secondary route of the

130

Creighton, O. H., Castles and Landscapes: Power, Community and Fortification in Medieval

England, (2002), p. 54

63

Slighe Mhór, linking Ballinasloe to Lough Croan, Roscommon and Boyle, eventually

linking it up to the other great road that served east and west in medieval Ireland, the

Slighe Assail.131

If we take the modern road linking Ballinasloe to Roscommon as the

basic successor to „Route 9‟, which, thanks to a comparison of maps, it invariably is,

then Dundonnell is less than 1km to the south east of the road.

Couple this with the fact that the Slighe Mhór itself is no more than 7km away from

Dundonnell, it being the most important road connecting Connacht with the east, gives

us some indication of the position that Dundonnell would have held if it was occupied

in the 13th

-century. The reason for its wider proximity to the Slighe Mhór can be

speculated further if we return to its postulated first phase and its position in relation to

the routeway. Stout noted that important secular ringforts were usually located away

from the main routes132

, presumably to lessen the impact that such an easily navigated

road would have in the practice of raiding, primarily of cattle.

The importance of holding these roads has been compared with having control of the

Shannon itself, and especially in these border areas, constant contesting of these

communication routeways would result in their changing hands regularly,133

something that does fit well with the annalistic evidence for castles in the area in the

mid-to-late 13th

-century.

131

Doran, “Medieval Communication Routes through Longford and Roscommon and Their Associated

Settlements”, p. 72, Table 1 132

Stout, The Irish Ringfort, p. 103, Fig. 29 133

Doran, “Medieval Communication Routes through Longford and Roscommon and Their Associated

Settlements”, p. 72-73

64

We know, from comparative research on the interaction of the Normans with Roman

roads in Britain, that English settlers would have had no problem with adopting the

pre-existing communications for their own uses, with, for instance, up to 80 per cent

of the castles of Hampshire being sited on or in the immediate vicinity of known

Roman roads.134

Therefore, to transfer the model to this island, the prospect of a 13th

-century ringwork

castle at Dundonnell on „Route 9‟ and in the close vicinity of the Slighe Mhór is not

beyond possibility. It may have been constructed there as a way of patrolling the vital

artery through an area that was both naturally treacherous135

and armed with hostile

locals. If this was the case, it was merely the continuity of a practice that had served

the English so well in the conquest of Britain. The importance of the road network to

this island in the medieval period has been largely overlooked by scholarship, with

notable exceptions, such as Ó Lochlainn,136

the first to treat it properly. The late

medieval and early modern history of this area shows quite well how important good

communication links were, due to a veritable absence of such links when the Irish

reacted to the re-colonisation of Connacht in the late 16th

century. (See 2.3) So for a

second phase site at Dundonnell to be placed where it was in relation to routeways,

shows the prior planning exhibited by the administrative powers in order to keep the

area under control.137

134

Creighton, Castles and Landscapes, p. 39, citing Hughes 1989:34 135

Doran twice refers to the significant hindrance and obstacle that bogland put on transport in this

period, therefore necessitating the important work of insuring the routeways remained free of raiders

and hostile locals. See Doran, “Medieval Communication Routes through Longford and Roscommon

and Their Associated Settlements”, pgs. 61 & 76 136

Ó Lochlainn, “Roadways in Ancient Ireland”, p. 465 137

A case could also be made for another road which connected Athlone to „Route 9‟ through the

bogland and an alternative route to Ballinasloe, which passes in very close vicinity to Dundonnell

65

Figure 15: Location of Dundonnell with regard to the major medieval routeways of the region

The link between castle and a religious community is something that has received

much study in recent times, and continental and British practice in this regard points

Castle, thus making the immediate vicinity of Dundonnell into a crossroads of sorts, around which the

modern day village, parish church and local amenities have all built up. The two medieval indicators of

settlement, the Carmelite cell at Bellaneeny, which is directly behind the modern church, and the castle,

the possible local administrative seat for the area, are both situated within a 4km² area of each other and

the modern village. Further research should be attempted here in order to see if there is a real link

between Dundonnell and a settlement that has developed to become the modern village.

66

very definitely towards a joining of histories, either by embracing an existing

institution or by founding one in conjunction with the castle, in order to serve the

population at the site and in any settlement that has grown up out of the castle.

Creighton devoted an entire chapter to the subject138

, while various other authors have

discussed this issue at length that are too numerous to name. 11th

and 12th

-century

earthwork castles in the Grand Caux region of Normandy show said castles on average

lying within 500m of a church, while a study of ten English counties show that over 40

castles in these areas had a church within 50m of the castle.139

We know that from 1181, English custom was applied to the Irish Church140

, merely

as another natural feature of the colonisation process. That the Irish Church was to

conduct itself along the lines of the Church of England and have its episcopate

anglicised, was just a way of keeping all bastions of authority singing from the same

hymn sheet, enabling a smoother settlement of the island.141

The construction of

fortifications next or near religious houses and churches was just one, relatively crude,

way of linking church and state, and the Irish examples don‟t disappoint either.142

To

take our own subject, we know that Bellaneeny had been in use by at least the 9th

century, and still in operation by 1567, when it was recorded in the Fiants of

Elizabeth.143

This shows that by 1236, when the Annals of Clonmacnoise assert that

138

“Power, Patronage and Parish: Castles and Ecclesiastical Landscapes”, in Creighton, Castles and

Landscapes, p. 110-132 139

Creighton, Castles and Landscapes, p. 111 140

Watt, J. A., The Church and the Two Nations in Medieval Ireland, (Cambridge: 1970), p. 46 141

Watt, The Church and the Two Nations in Medieval Ireland, p. 83 142

Although castle and church are not as readily associated in Ireland as in Britain, examples show that

the practice was quite pronounced nonetheless. Castlerahan, Co. Cavan, Kilpipe, Co. Wicklow,

Rathangan, Co. Kildare, Milltown, Co. Meath, Cloncurry, Co. Kildare & Donohill motte, Moat quarter,

Co. Tipperary all show evidence of being closely located to medieval churches. See Sweetman, The

Medieval Castles of Ireland, Figs. 3,4,12, 18, 21, pgs. 8-31 for examples. 143

Egan, “The Carmelite Cell of Bealaneny”, pgs. 19-20

67

Onagh was built, Bellaneeny was in full operation and, as such, suggests that any 13th

-

century fortification to have been built at Dundonnell could have had the proximity to

the religious centre in mind also.

To go back briefly and deal with the relationship between ecclesiastical sites and

ringforts, where there is a possibility that the first phase of Dundonnell may have had

interaction with Bellaneeny. Stout notes that ringforts in the immediate environment of

churches and ecclesiastical sites generally points to patronage and/or lay communities

congregating around the site144

, but this is as detailed as we can get on this theory.

Having dealt with the possibility that a 13th

-century ringwork castle was constructed

over an earlier ringfort, we must now attempt and understand why it is sited

specifically over said structure. Having dealt already with the physical attributes that

building in that area provided; the water supply, and the proximity to pre-existing

communication routes and ecclesiastic centres, we must now try and understand why

the site itself, our first phase postulated ringfort, may have been chosen. Now,

obviously, at this point, we can only theorise, but with the backing of circumstantial

evidence, we can make some realistic claims.

Firstly, we must make a nod towards the practicality of building over existing sites.

The presence of existing fortifications reduced greatly the amount of time and labour

required in order to complete an earthwork, whether it was aided by a natural

144

Stout, The Irish Ringfort, pgs. 100-106

68

escarpment or feature, or a pre-existing constructed defence.145

Old Sarum in Wiltshire

and Mount Caburn in Glynde, East Sussex are just two English examples, while at

Dunsilly motte, Co. Antrim, and Lismahon and Rathmullan mottes, Co. Down, show

that the practice continued and adapted to the ringfort, for practical reasons at least.146

Cóilín Ó Drisceoil brings this argument even further when he asserts that eight of the

fifteen motte castles excavated in Ireland in the modern period have ringfort

precursors,147

a significant number, even if the sample size is pitifully small. The fact

also that great care was taken by castle builders in Ireland to construct over pre-

existing royal sites tells its own tale as regards their motives and how they sought to

pacify the local area.148

The symbolic or status-based relationship that this site could have had to the

surrounding area may also have been a reason for it being constructed over.149

Creighton discusses the re-occupation of Saxon sites with regal, high-status or

religious connotations in the Norman Conquest of England in this regard. He pointed

to a large and ever increasing corpus of archaeological evidence of Saxon élite

residences being expropriated and remodelled to suit a new Norman landlord.150

And

while he does insert the required provisos to ensure that we don‟t get carried away

with this idea, his image of what this could do to the local population is striking -

145

Creighton, Castles and Landscapes, p. 37-38 & 69 146

Sweetman, D., Irish Castles and Fortified Houses, (Dublin: 1995), p. 6 147

Ó Drisceoil, C., “Recycled Ringforts: The Evidence from Archaeological Excavation for the

Conversion of Pre-Existing Monuments to Motte Castles in Medieval Ireland”, Journal of the County

Louth Archaeological and Historical Society, Vol. 25, No. 2 (2002), pp. 189-201 148

Herity, M., “Motes and Mounds at Royal Sites in Ireland”, The Journal of the Royal Society of

Antiquaries of Ireland, Vol. 123 (1993), pp. 127-151 149

Here we must recall that the defence of a ringfort relied much more heavily on communication links

with the settlements surrounding it, than the physical defences of the homestead itself, along with the

power and prestige of the chief occupant being used to deter would be assailants from attack and

raiding. 150

Creighton, Castles and Landscapes, pgs. 70-71

69

“Saxon thegns tended to have expansive residences, and their appropriation by new

Norman landlords often meant their remodelling as more compact, heavily defended

and visually imposing units that presumably had a strong local psychological impact,

especially where a new motte towered over the tenants.”151

As with many of the features that characterised the Norman Conquest of Britain, they

can be transposed upon the English colonisation of Ireland, just over one century later,

with relative ease, increasing the possibility of Dundonnell having two phases by the

end of the 13th

-century.

Whether these ringforts were still inhabited and therefore held a level of prestige for

the occupants has to be taken in a case-by-case manner, and a new site may well have

been built over a long deserted ringfort. However, we must keep in mind that no

evidence exists that shows there may have been a widespread desertion of the ringfort

post 1169, and ringforts were still very much in use in the medieval, and maybe even

into the early modern period in some cases.152

Therefore, the likelihood of re-occupying an, until recently, inhabited site, be it

acquired pacify, conformingly, or through the force of arms, cannot be discounted

with regard to 13th

-century Dundonnell. As a result, the idea of assuming residence of

what may have been a powerful local chieftain, albeit something we cannot prove,

nevertheless, isn‟t out of the question.

151

Creighton, Castles and Landscapes, p. 71, citing Hinton, D. A., Archaeology, Economy and Society:

England from the Fifth to the Fifteenth Century, (London: 1990), p. 110 152

Barrett, G. F., Graham, B. J., Lynn, C. J., “Some Considerations concerning the Dating and

Distribution of Ring-Forts in Ireland”, Ulster Journal of Archaeology, Third Series, Vol. 38 (1975), pp.

33-47, p. 37

70

4.3 - Defensibility of the site

Now we must inevitably look at the defensibility of the site, in order to understand the

role of a castle in its traditional capacity. Here we come across a difficult balancing of

priorities with regard to the true military value that a castle brings to a region. There is

a constantly evolving argument with regard to the subject of a castle‟s defensibility,

and the position that defence holds in the list of attributes that come with the building

of same. If we look at the English examples of castles of even earlier construction than

the postulated ringwork at Dundonnell, such as the ones studied by Liddiard153

, we

find that the practical military features of some sites have become purely decorative,

leaving blind spots and other such problems, should the fortification have been

seriously attacked.154

On face value, this situation could also be seen with regard to places such as the 13th

-

century phase of Roscommon Castle, with its location choice traditionally being called

into question155

, along with the practicality of the fortifications themselves. However,

the political situation that Liddiard found in Norfolk was vastly different from the

situation that was to be played out in 13th

-century Roscommon, and a fresh look by

Murphy on the subject of Roscommon Castle has yielded much more fruitful, and in

the context, practically-based conclusions.156

153

Liddiard, R., “Landscapes of Lordship: Norman Castles and the Countryside in Medieval Norfolk,

1066-1200”, BAR British Series 309 (2000) 154

Liddiard, “Landscapes of Lordship”, p. 3 155

This argument goes as far back as Weld, I., Statistical Survey of the County of Roscommon, drawn

up under the directions of the Royal Dublin Society, (Dublin: 1832), p. 393 156

Murphy, M. “Roscommon Castle: Underestimated in Terms of Location?”, Journal of the Galway

Archaeological and Historical Society, Vol. 55 (2003), pp. 38-49

71

Therefore it is with these provisos in mind that we must carefully analyse the

defensibility of Dundonnell, primarily in a 13th

-century context, but this will suffice

also for the definite 16th

-to-17th

century 3rd

phase.157

The site is located on a lowland

stretch that is served to the south by an area of high ground. To the traditionalist, this

could be seen as compromising its defensibility due to a possible artillery or projectile

attack. However, this argument falls down as soon as the theoretical castle in question

is approximately two hundred metres away from the potentially dangerous high

ground.158

Range is the key factor, and with the high ground to the south of

Dundonnell petering out on average over five hundred metres before we reach the site,

the advantage of the high ground is effectively nullified. In addition to this, the First

Edition 6 inch Maps show that the earthwork is located (see 3.1) on what was a dry

promontory surrounded by bogland, meaning that it was only approached safely from

the north. The choice of the site here, along with the careful positioning of the entrance

to the north-east of the site, lessens any advantage that attacking from the southern

high ground could have had, if one were to be brave enough to traverse the now

disappeared bogland in the process.

This shows that the most obvious locations for a castle in the eyes of the modern

observer might not necessarily make it the best suited for the builders at the time. The

lower ground provided better access to water than the high ground; it was much closer

to the road that we can presume it sought to control, and the religious centre also.

157

Because, although there are, in essence, 200+ years of separation between the 2nd and 3rd phases,

the key aspects that made the site viable for defence, along with the other reasons that it was chosen for

in the 13th-century, remained relatively unchanged, despite the advance to firearms. 158

Liddiard, “Landscapes of Lordship”, pgs. 6-7

72

These factors show that defence was clearly not the sole preoccupation of the builders

of this fortification, pointing to the wider social and societal function that Dundonnell

could have served for this presumed period of its habitation. Creighton shows that

castles should not be pigeon-holed as strictly as their appearance would suggest, and

the more functions we can ascribe for a castle can only help to elaborate why a

location that would not seem easily defendable could still be chosen.159

We have also

seen that only through proper inspection of the landscape, not just as it is today, but

also exploring the possibility that these lands were modified over time by agriculture

and land reclamation, do we truly understand the factors pondered when it came to

constructing any of the three phases at the site of Dundonnell.

4.4 - Conclusions

To conclude the siting chapter, I would like to highlight a very important point made by

Lilliard in relation to the defensibility of a site, something that is often overlooked

when archaeological discussions concerning defence and siting is put forward. This is

the essential human element. If a fortification needs to be defended, be it a concentric

castle or a simple timber tower within an earthwork, it is as much about the conviction

and application of the defenders as it is about the materials and siting of the castle, that

decides the outcome.160

To recap, I have shown the aspects of the wider landscape and the possible pre-existing

elements that are present that could have served Dundonnell. The builders of each of

the phases had their own motives for constructing at Dundonnell, features that go

159

Creighton, Castles and Landscapes, p. 36 160

Lilliard, “Landscapes of Lordship”, p. 6

73

beyond the practical militaristic concerns. This has been undertaken in order to help in

analysing the site and will thus become of further use to this thesis presently.

74

Chapter 5 - Phase 1 at Dundonnell: the postulated Ringfort phase

5.0 - Introduction

The aim of this chapter is to explore the possibility that the earliest phase of habitation

at Dundonnell was that of an early medieval ringfort.

5.1 - The postulated ringfort at Dundonnell

The first studies to suggest that Dundonnell had at its origins as an early medieval

ringfort were the two Graham articles, written in 1988, “Medieval Settlement in

County Roscommon”161

and “Medieval Timber and Earthwork Fortifications in

Western Ireland”162

. O‟Keeffe follows these up, with his 1998 article “The

Fortifications of Western Ireland, AD 1100-1300, and Their Interpretation”.163

O‟Conor‟s 2007 article, “English settlement and change in Roscommon during the late

sixteenth and seventeenth centuries”164

, is the most recent to date to describe the site as

having a ringfort as its first phase. Therefore, why do we think that there was a ringfort

at the site of Dundonnell? The problem here is that none of the above scholars has

really explained why they believe that there was a ringfort here. Therefore, in order to

continue, I will once again discuss the evidence that I believe points to Dundonnell

having a 1st phase ringfort.

The first piece of evidence that points to Dundonnell having a ringfort is the

morphology of its earthwork. (Fig. 8) It is an oval, bivallate constructed earthwork,

161

Graham, B. J., “Medieval Settlement in County Roscommon”, pgs. 28-29 162

Graham, B. J., “Medieval Timber and Earthwork Fortifications in Western Ireland”, pgs. 122-123 163

O‟Keeffe, T., “The Fortifications of Western Ireland, AD 1100-1300, and Their Interpretation”,

p.190 164

O‟Conor, K., “English settlement and change in Roscommon during the late sixteenth and

seventeenth centuries”, pgs. 191-192

75

with a north-eastern facing entrance. (See 3.2a for full description) The morphology of

this site is highly indicative of the type of features associated with a large ringfort, the

bivallate bank and ditch system providing extra protection to what may have been a

relatively high-status dwelling. The north-eastern facing entrance, if it corresponds

with the original entrance, faced in the general direction of the morning sun, indicating

the probability of the site being used for human habitation.165

There is also the

likelihood that the entrance could have been preferably more easterly in position, if it

wasn‟t for the fact that it would probably result in the entrance having to cross bogland

in the process. (See 3.1) However, the formidable nature of the ditch and banks

suggests that at some point this site was refortified, something which will be discussed

later. Overall, however, the morphology the earthwork at Dundonnell does show

positive evidence for a ringfort.

The proximity of Dundonnell to a number of other ringforts, which would have

provided protection, its abundant water supply, its position adjacent to a prominent

local religious centre and its location away from the main, and therefore potentially

dangerous, routeway of the area, the Slighe Mhór, all show signs that a ringfort could

have been located at Dundonnell. (See 4.1) Although these factors are only

circumstantial in nature, when combined with the morphological evidence, it does aid

in the case for Dundonnell having a ringfort first phase.

Finally, the placename evidence provides us with an interesting source of information

with which to view Dundonnell. “Placenames as Indicators of Settlement”, published

165

Stout, The Irish Ringfort, pgs. 18-19

76

by A. S. Mac Shamhráin in 1991, provides a basis of Irish placename terminology, and

includes both dúnand lios as indicators of secular, earthen sites.166

This form of

evidence was incorporated into the work of both Graham and O‟Keeffe, resulted in

somewhat conflicting conclusions. Graham cited Orpen, who showed that the likes of

dún or liss could be added to a townland name after the erection of Norman

fortifications167

. Conversely, O‟Keeffe used the dún element of a place name to show

its Irish, pre-Norman origins,168

but not specifically with regard to Dundonnell itself.

Either way, the placename evidence does indicate quite clearly that a fortification of

some description existed at Dundonnell at an earlier period than the stronghouse.

The townland name of Dundonnell was first recorded as in 1420, with its mention in

the Registry of Clonmacnoise.169

This late date, along with its position adjacent to

Onagh townland (Fig. 12), means we cannot rule out the earlier positioning of the two

areas within an older, larger townland, something that is not without precedent in

Taughmaconnell, as seen with the 1636 Stafford Survey reports.170

This in itself

allows for the possibility that Dundonnell fell under the older townland name of

Onagh in the 13th

century sources, and its subsequent change in placename to

Dundonnell.

166

Mac Shamhráin, A.S., “Placenames as Indicators of Settlement”, Archaeology Ireland, Vol. 5, No. 3

(Autumn 1991), pp. 19-21, pgs. 20-21 167

Orpen, G. H., “The mote of Lisardowlan, County Longford”, The Journal of the Royal Society of

Antiquaries of Ireland, Fifth Series, Vol. 40, No. 3, [Fifth Series, Vol. 20], (Sep. 30, 1910), pp. 223-225,

p. 225 168

O‟Keeffe, “The Fortifications of Western Ireland, AD 1100-1300, and Their Interpretation”, p. 188 169

O‟Donovan, J., “The Registry of Clonmacnoise: With Notes and Introductory Remarks”, The Journal

of the Kilkenny and South-East of Ireland Archaeological Society, New Series, Vol. 1, No. 2 (1857),

pp.444-460, p. 455, referring to the donation of lands in Moyfinn to the monastery in the 6th

century. 170

Taughmaconnell: a history, p. 13

77

Figure 16: Ordnance Survey Map containing the extents of Dundonnell and Onagh townlands

5.2 - The ringfort in Ireland

5.2a - Number & Distribution

The ringfort is the most numerous but yet one of the most mysterious monument types

present on the Irish landscape today. The broad consensus holds that there are

approximately 40,000 remaining ringforts on the island today, with the number

possibly increased to 50,000 if we take into account the ones that have been removed

from the landscape due to modern land use and cultivation.171

The distribution of the

ringfort varies from region to region with density particularly high in South-East

Connacht, where Dundonnell is situated.172

The ringfort distribution for the area

around Dundonnell conforms to the statistics, with the total of six ringforts present in a

9km² area around our site.

5.2b - Morphology

171

Ó Corráin, D., “Ireland c.800: aspects of society”, in Ó Cróinín, D., (ed.), The New History of Ireland

1: Prehistoric and Early Ireland, (Oxford: 2008), pp. 549-608, p. 550 is merely one example of the

multitudes of studies that accept this figure. 172

Stout, M., The Irish Ringfort, p. 81-82 & Fig. 14, p. 60

78

Considering the huge corpus of examples and wide range of both dates and types of

ringfort in the Irish landscape, it is notoriously difficult to correctly provide a

morphological definition of the monument type.173

In order to provide the most

complete definition for the ringfort, therefore, I believe simplicity is best and refer to

the definition provided by Ó Ríordáin, largely accepted as the most accurate

description:

“In its simplest form the ringfort may be described as a space most frequently circular,

surrounded by a bank and fosse or simply by a rampart of stone. The bank is generally

built by piling up inside the fosse the material obtained by digging the latter. Ringforts

vary very considerably in size. In the more elaborately defended examples, the

defences take up a much greater area than that of the enclosure.”174

To apply anything more detailed or nuanced to the morphology would only serve to

limit the definition and, thus, the understanding of the site type. The only addition I

believe is needed here is to highlight that the range in complexity can often tell us a

great deal about the function behind any given ringfort, and should always be explored

in full.

5.2c - Function

Its primary functions seem to centre around habitation and defence and, in some cases,

prestige, which in the highly militaristic world of early medieval Ireland, are roles that

are intrinsically linked. The ditch and bank system is something that is evident in the

173

Edwards, N., The Archaeology of Early Medieval Ireland, (London: 1990), pgs. 12-14 174

Ó Ríordáin, S., Antiquities of the Irish countryside, 5th ed. (London: 1979), pgs. 29-30

79

vast majority of ringforts, with this feature obviously increasing the defensibility of a

site the deeper it was. The circular or oval shape of the site was also constructed in

order to lessen or eradicate the possibilities of blind spots in the defences in the case of

attack, something that moated sites were prone to upon their introduction.175

The

combination of the two, along with the likelihood of a fence or palisade to run along

the top of the bank176

, if not a sod, cob or stone defensive wall, show that these sites

espoused at least a minimal level of protection for the people, livestock, and

possessions within. Coupling this with the previously mentioned overlapping visual

defence that seemed even more important than the physical defences of the site, show

that serious considerations were as much given to location in relation to other sites as

to the depth of the ditch.

The fact the entrance faced east or south-east in over half of ringforts recorded on the

island also strongly suggests that many ringforts were much more than livestock

enclosures, and permanent habitation took place in many instances.177

With a

functionality range as wide as that of the ringfort, there are inevitably going to be

problems in successfully determining a conclusive role for the ringfort in society. The

problem is exacerbated when scholars fall into the easily trod route of generalising a

monument type due to the huge quantity apparent on the landscape. Therefore, I

believe that the only true way to study a ringfort is through case-by-case study, in

175

Stout, The Irish Ringfort, p. 15 176

Although Mallory and McNeill have stated that there were no palisades found on the ringforts in

their study, and used it as one of the five reasons why a ringfort wasn‟t realistically defensible. See

Mallory, J. & McNeill, T., The archaeology of Ulster, (Belfast: 1991), pgs. 196-199. However there is

reason to believe that palisades or timber barriers of a kind may have sat on top of a bank in some

instances, much like a wattle and daub wall, thus leaving little, if any, evidence on the archaeological

record. This is also besides the fact that there is a large body of evidence for timber palisades on sites

outside their study area. 177

Stout, The Irish Ringfort, pgs. 18-19

80

order to avoid generalisations that can only harm the field of research in the future. I

believe a similar approach should be taken on ringforts, despite the numbers, as is

currently occurring with castle studies by the likes of Creighton,178

espousing a more

thorough and analytical approach. In this way, generalisations such as the

conventional estimations of the ringfort being a poor defensive site can be explained

properly, with a view to the societal goals of the period. Factors such as the wealth

associated with cattle and cattle raiding and the lack of primogeniture show us the

reasons behind the construction of ringforts in Ireland over a more physically

imposing, and expensive site. Instead the preference of defence in depth was just one

such approach employed in Ireland.179

The problem for us, however, is that defence

like this is nigh on impossible to detect in the archaeological record.

5.2d - Dating

Dating, along with much of the other facets of ringfort study, is fraught with difficulty.

While its origins as a monument type has been dated back as far as the late Neolithic

and early Bronze Age in some cases180

, scholars have also argued that the ringfort was

still in use, if not being still constructed, into the 12th

and 13th

-centuries.181

So where

does one draw the line? The safest parameters with which to date the ringfort, in my

eyes, is in the period between the 5th

-century AD and the late 12th

to early 13th

-

178

Creighton, Castles and Landscapes: Power, Community and Fortification in Medieval England, p.

36 179

Stout, The Irish Ringfort, p. 20 180

Edwards, The Archaeology of Early Medieval Ireland, p. 17 181

Barrett, G. F., Graham, B. J., Lynn, C. J., “Some Considerations concerning the Dating and

Distribution of Ring-Forts in Ireland”, Ulster Journal of Archaeology, Third Series, Vol. 38 (1975), pp.

33-47, p. 35

81

centuries. Of course this can only be augmented through further research into what is,

an admittedly a small excavated sample size.182

5.3 - Conclusions

Dundonnell may have been a ringfort in its 1st phase; only excavation will ever

conclusively tell if this is the case. As a result, this has to be one of the main aims of

any future study that is ever undertaken at Dundonnell. As noted previously, (see 4.2)

colonisers, originally the Normans, often made ready use of older sites, which in the

Irish examples, were predominantly ringforts.183

Therefore, a 13th

-century castle being

placed on top of a pre-existing ringfort was part of the common practice of the English

in the 12th

and 13th

-centuries, and as a result, should be taken into account in the case

of Dundonnell also. The function of Dundonnell in the Early Medieval period may

have stretched to it being the residence of a wealthy member of the Uí Maine túath in

the strategically significant area of South Roscommon in the period before the arrival

of the English.

182

Stout, The Irish Ringfort, Fig. 2, p. 29 183

Sweetman, D., Irish Castles and Fortified Houses, (Dublin: 1995), p. 6 & Ó Drisceoil, C., “Recycled

Ringforts: The Evidence from Archaeological Excavation for the Conversion of Pre-Existing

Monuments to Motte Castles in Medieval Ireland”, Journal of the County Louth Archaeological and

Historical Society, Vol. 25, No. 2 (2002), pp. 189-201

82

Chapter 6 - Phase 2 at Dundonnell: The postulated Ringwork

6.0 - Introduction

The aim of this chapter is to explore the possibility that the second phase of habitation

at Dundonnell was a 13th

-century ringwork castle built by the English in 1236.

6.1 - The postulated ringwork at Dundonnell

The first advocate for Dundonnell possibly having a second phase ringwork was

Graham, in his two 1988 articles.184

Barry, in The Archaeology of Medieval Ireland,

echoed this thought, where he consulted Graham in relation to his labelling of

Dundonnell as a possible ringwork.185

While there have been detractors this theory in

more recent times186

, along with a suggestion made for an alternative location by a

local historian in 1974187

, Linda Doran is the most recent scholar to suggest that

Dundonnell could have had a ringwork phase in her 2004 article “Medieval

Communication Routes through Longford and Roscommon and Their Associated

Settlements”.188

The lively debate about the location of Onagh and the possibility of

Dundonnell being occupied in the 13th

-century shows just how divisive the point is.

Why, therefore, do I believe there was a ringwork at Dundonnell?

184

Graham, B., “Medieval Timber and Earthwork Fortifications in Western Ireland”, pgs. 122-123; &

Graham, B., “Medieval Settlement in County Roscommon”, pgs. 28-29 185

Barry, T., The Archaeology of Medieval Ireland, pgs. 52-53 186

O‟Conor, K., “English settlement and change in Roscommon during the late sixteenth and seventeenth

centuries”, pgs. 191-192; & O‟Keefe, T., “The Fortifications of Western Ireland, AD 1100-1300, and

Their Interpretation”, p. 190 and endnote 54, p. 198 187

Moore, Dr. A., “Mac Keoghs of Moyfin”, Old Athlone Society Journal, Vol. I, No. 4, (1974-75), pp.

234-237, p. 234. The problem with Moore‟s suggestion that it had no evidence to back up his choice as

being Onagh, along with the site in question being little more that a ringfort or ringbarrow on the

physical landscape. 188

Doran, L., “Medieval Communication Routes through Longford and Roscommon and Their

Associated Settlements”.

83

In order to prove the likelihood of Dundonnell having a ringwork second phase in its

history, three key issues must be dealt with - the morphology of the site, its location in

the wider landscape, and the association of Dundonnell with the castle of Onagh in the

historical sources.

The morphology of the site at Dundonnell, with reference to its earthworks

specifically, holds an oval enclosure that is surrounded on all sides by a system of

deep ditches and accompanying banks, that has been argued (See 5.1 & 3.2a) to be the

remnants of an early medieval ringfort of indeterminate date. However, it seems, due

to the extent of the defences, any ringfort on the site must have been refortified in

order to facilitate future habitation on it. This means that while it could have been

refortified to serve the masonry building dated to the 16th

and 17th

-centuries, it may

also be part of a ringwork castle, built sometime in the 13th

-century. If we analyse the

evidence that is available to us, a realistic claim can be made for Dundonnell having a

13th

-century ringwork as its second phase.

With regard to its location, Dundonnell is situated in close vicinity to a steady water

supply and in direct proximity to two vital land routeways, „Route 9‟ and the Slighe

Mhór. These essential characteristics of the location are supplemented by the potential

importance of having a local religious centre in the immediate area. These are all

factors that English settlers seek to utilise when they are determining a location for

their fortified residence and administrative centre. If we analyse the idea of building

over a pre-existing site with regard to English practice, the multitude of both Irish and

British examples of this adoption does allow us to postulate the likelihood of this

84

happening at Dundonnell also. This was done for both practical and symbolic reasons,

and either theory is valid in the case of Dundonnell (for full discussion see 4.2)

The historical references (see 2.5) to a castle being built at Onagh in 1236 provide us

simultaneously with a possible answer and a problem in relation to Dundonnell. If we

wish to associate Onagh with the postulated 13th

-century ringwork castle at

Dundonnell, we must deal with the fact that Onagh is a townland separate from

Dundonnell in itself. However, the lack of any feature of equivalent size and

complexity in the townland of Onagh, coupled with the two townlands being adjacent,

do pose some interesting points as to Onagh‟s location. The first reference for the

townland name of Dundonnell only surfaces in the 15th

-century189

, pointing towards a

possible amalgamation of the two townlands at an earlier date, when the castle of

Onagh was built. If we analyse the site type as being of ringwork/refortified ringfort

construction, it does fit with the historical accounts that state that eastern Connacht

was in a state of flux in the 1200s. A ringwork castle, built on a pre-existing defended

site190

, would be the most rapidly constructed fortification191

, thus facilitating the need

to swiftly control a hostile territory. Combined, they provide us with a viable location

for the castle of Onagh being that of Dundonnell on the physical plane. However, this

can only ever be verified conclusively through excavation.

189

O‟Donovan, J., “The Registry of Clonmacnoise: With Notes and Introductory Remarks”, p. 455 190

Sweetman, Irish Castles and Fortified Houses, p. 6, with Kenyon suggesting that the figure of

castles, mottes in his case, in Ireland being built over pre-existing fortifications could reach the 400s,

something, however, that could only be verified only with excavation. See Kenyon, Medieval

Fortifications, p. 7 191

De Meulemeester, J., O‟Conor, K., “Chapter 11: Fortifications”, in The Archaeology of Medieval

Europe: Vol. – Eighth to Twelfth Centuries AD, (Aarhus: 2007), pp. 316-341, pgs. 327-328

85

6.2 -The ringwork in Ireland

6.2a - Number and Distribution

The numbers of ringwork castles in the Irish landscape has been the source of much

debate for archaeologists in the modern era, due to a large degree to the features that

would have it readily confused with the ringfort. Its construction often lends itself to

being mistaken for a ringfort, and vice-versa.192

However, for the purposes of this

work, the numbers and distribution map provided by Barry in The Archaeology of

Medieval Ireland provides the vast majority of identified ringwork sites on the

island.193

The number of ringwork castles stretches to approximately fifty on the Irish

landscape to date, but it is their distribution that is most interesting. At least ten

survive in the modern county of Wexford, which is where the English first arrived in

1169. Other areas of relatively high distribution occur in Roscommon and North

Tipperary. This will be explained in more detail later in relation to the function of a

ringwork castle.

6.2b - Morphology

The position of the ringwork castle in the Irish archaeological landscape is an

intriguing and, often, difficult to analyse feature. The fact that the majority of modern

castle scholarship simplifies the study of earthworks into two varieties is, in itself,

rather archaic, and some scholars have deliberately sought to distance their work from

this clear-cut and, ultimately, misleading division.194

If we are to assign a castle a

certain type therefore, we must proceed with that label rather cautiously, as to over-

192

Barry, T., The Archaeology of Medieval Ireland, p. 45 193

Barry, The Archaeology of Medieval Ireland, Fig. 14, pgs. 52-53 194

One such revisionist study occurs with Higham, R. A., “Early Castles in Devon, 1068-1201”,

Château Gaillard 9-10, (1982), pp. 102-115

86

generalise the monument would be an unfortunate disservice to it and the discipline

itself. With this in mind, a fair definition of what a ringwork is, must be provided, with

Sweetman describing it as such:

“In England and Wales it (the ringwork) is, in its simplest form, an area enclosed by a

fosse and rampart. It has also been defined as having a minimum height of 2m above

the level of the outside defences with the enclosed area disproportionately small

compared to the massive enclosing elements. However, in Ireland we can expand on

these definitions by saying the bank(s) are more pronounced and the fosse is wider

than one would expect to find on a ringfort. The entrance to the ringwork is also

distinguishable from a ringfort in that it will often have a pronounced ramp and each

side of the gap in the rampart will be faced with stone.”195

Without generalising, there are two separate earthwork types evident on a Normanised

landscape, be it in Ireland or Britain. Some scholars have attempted to provide clarity

to situations where the monument in question has features of both castle types,

leading, unfortunately to such oddities as the „ring-motte‟.196

Instead, scholars are

better served if they study each individual monument on their merits, assigning the

earthwork type, be it motte or ringwork, as a broad umbrella with which to work from.

Realistically, function defines the morphology of any site, and over-categorisation

should be avoided.

195

Sweetman, The Medieval Castles of Ireland, p. 4, citing King, D. & Alcock, L., “The Ringworks of

England and Wales” in Château Gaillard 3, (1969), pp. 90-127 within. 196

Kenyon, J. R., Medieval Fortifications, (Leicester: 1996), p. 5

87

6.2c - Defences

The ringwork works as a castle type based primarily on a need to serve practical

issues. The motte, as seen with the longevity of construction needed to raise the large

earthwork, has much of its motives based around intimidating the surrounding

countryside, along with acting as a sign of prestige on the part of the owner of the

residence. On the other hand, the ringwork derives most of its defensive prowess from

the peripheral palisade, coupled with a ditch and bank system, along with a fortified

gate tower, or tower on the internal summit.197

6.2d - Function

There have been many attempts to provide concrete reasons why one castle type was

preferred over another, based on the morphology of the site type, its date range, its

speed of construction etc. These attempts seem to work on the grand scale, but don‟t

serve in the case of every individual site. As a result, conclusions drawn by more

recent research has accepted that the construction of one castle type over another

seems to be based primarily on personal preference.198

To take one example of

preference being at the heart of the decision, we see that ringworks were constructed in

south Glamorgan in a large number due primarily to the nature of the soil and

geology.199

Therefore, we can really only look at what makes one choose the ringwork

over the motte, in order to understand its function properly.

197

Barry, The Archaeology of Medieval Ireland, p. 45 198

Kenyon, Medieval Fortifications, p. 5 199

Spurgeon, C.J., “The castles of Glamorgan: some sites and theories of general interest”, Château

Gaillard 13, (1978b), pp. 203-226, pgs. 206-207

88

One of the advantages of constructing a ringwork deals primarily with the speed of

construction of such a site200

, thus leading to the belief that the ringwork may have

been chosen in some instances as a „campaign castle‟. This theory has substantial

evidence to back it up, given that many stone castles have been built over ringwork

precursors, such as Castle Neroche, Somerset, and the Tower of London201

in Britain,

and the castles of Clonmacnoise and Trim in Ireland. Coupling this with

aforementioned examples from Wexford202

, the landing point for English troops in

1169, and we get a clear idea of one of the reasons why a relatively low structure with

a substantial ditch and bank system and timber gate tower was chosen in these cases.

However, that isn‟t to say that these ringworks could not act in any capacity beyond

the realm of campaign and invasion. Ballysimon ringwork, Co. Limerick is one

example of many, of a permanently inhabited ringwork, which doesn‟t follow the

usual model of a „campaign castle‟.203

Initially, the ringwork acted as a way of quickly claiming the land, and as a practical

base from which to go forth into the surrounding landscape. The ringwork phase of the

Tower of London is a prime example of an early attempt to place a military

fortification in the largest settlement in Saxon England and thus start taking over

proceedings after Hastings. On a more local level, the ringwork phase at

Clonmacnoise was an attempt to control the power of the economic heartland of the

midlands with its artery on the Shannon. But as seen with the masonry castles that

200

Kenyon, Medieval Fortifications, p. 7 201

Kenyon, Medieval Fortifications, p. 7 202

Barry, The Archaeology of Medieval Ireland, p. 52-53, Fig. 14 203

Collins, T., & Cummins, A., “Excavation of a Medieval Ringwork at Ballysimon County Limerick”,

Aegis Archaeology Reports 1 (2001), pgs. 11-12

89

replaced them, their role in the region was deemed established enough to put the stone

castle on the site thereafter.

6.2e - Dating

The date ranges that apply for the ringwork castle in Ireland effectively begin in the

mid 12th

-century and end, depending on the area, by about the end of the 13th

-century,

when the masonry castle has already been employed on a large scale, however there

are some examples in Ireland as late as the 14th

-century.204

6.3 - Conclusions

Dundonnell may have been the royal castle of Onagh.205

The postulated ringwork

would have had serious timber defences to supplement the elaborate ditch and bank

system that surrounds the enclosure. The interior probably would have been full of

administrative and domestic buildings of timber and cob construction, along with a

heavily defended gatehouse, or more likely in this case, given the size of the

entranceway, a timber tower on its summit. The function of Onagh was to protect

settlers in the area of South Roscommon along with protecting the important and

economically profitable routeways that came through the South Roscommon area, the

cantred of Uí Maine. The pomp and imposing nature of a motte summit with a timber

tower could not be entertained for Dundonnell, as the pressing nature of the political

scene206

meant that practicality and speed won out over psychology. This is not to say

that the ringwork at Dundonnell was not intended for permanent habitation either, as

204

Higham, R., Barker, P., Timber Castles, (London: 1992), pgs. 67 & 83-88 205

Calendar of documents relating to Ireland, p. 417, entry 2792, year 1245 206

Annals of Clonmacnoise, p. 235

90

the parallel example of Ballysimon and the longevity of settlement at Onagh, at least,

are testament to.

91

Chapter 7 - The Third Phase at Dundonnell: The Stronghouse

7.0 - Introduction

The aim of this chapter is to analyse physical remains of the postulated third phase of

habitation at Dundonnell, the proposed 16th

or 17th

-century masonry stronghouse.

7.1 - The stronghouse at Dundonnell

What date is the stronghouse? Archaeologist and architectural historians often use

window types and different types of stone dressing to date medieval and post-medieval

buildings.207

Unfortunately no datable windows can be seen with the building.

However, some of the stones used in the structure are roughly punch dressed. This

form of dressing stone in Ireland dates from c. 1400 into the 17th

-century.208

This date

is, however, obviously too broad for our purposes. At least nine definite gunloops can

be seen in the building today. Gunloops can be seen in later tower houses of late 16th

and early-to-mid 17th

-century date.209

The earliest gunloop known in Roscommon

comes from Roscommon Castle, 27 kilometres north of our site and it dates to the

early 1580s.210

Furthermore, the diagonal-shaped chimneystacks seen on each of the

gable walls are important also. These occur on buildings, usually tower houses and

fortified houses of late 16th

and early-to-mid 17th

-century date.211

In architectural

terms, this all suggests that the stronghouse at Dundonnell is of late 16th

or early-to-

mid 17th

-century in date. The scholarship to date has been broad in its interpretation of

the date and function of the masonry phase of the site, ranging from it being proposed

207

Leask, H. G., Irish Castles and Castellated Houses, (Dundalk: 1941), pgs. 23-24, Fig. 14 208

McAffe, P., Irish Stone Walls, (Dublin: 1997) 209

Sweetman, D., The Medieval Castles of Ireland, (Cork: 1999), pgs. 148, 159 & 164, for example 210

Murphy, M. & O‟Conor, K., Roscommon Castle: A Visitor’s Guide, (Roscommon: 2008), p. 30 211

Leask, Irish Castles and Castellated Houses, Fig. 55, p. 90, of Derryhivenny Castle, Co. Galway circa

1643, & p. 104

92

as having been identified as Crown property, repossessed, then used as an English

garrison post in Roscommon during the Nine Year‟s War212

at the turn of the 17th

-

century, to its possible capacity as a Mac Keogh residence in the mid-17th

-century.213

The case can be made for both scenarios, or even the possibility of it occupied on both

occasions, something that we can only speculate on presently.

7.2 – Stronghouse or Fortified House?

The case for Dundonnell being classified as a stronghouse rather than a fortified

house,214

lies primarily on the architectural evidence, as described in chapter 3 (See

3.2b). The practical nature of the building, with its myriad of gun loops, small

windows, deep walls, and served by a machicolation strategically placed over the

entranceway, point to a class of building that doesn‟t correspond neatly with the near

contemporary fortified house. Whereas the fortified house was another step in the

direction towards what would become the undefended Elizabethan mansion of the late

17th

-century215

, the stronghouse was an entirely different creature. Dundonnell, with its

lack of aesthetic based architecture, features such as size, symmetry, high gables and

transomed and mullioned windows, instead makes heavy use of defence, thus inferring

that its function was much less residence-based, much more a military outpost in a

hostile territory.

212

Cronin, T., “The Elizabethan colony in Co. Roscommon”, pp. 107-120, in Murtagh, H. (ed.), Irish

midlands studies-essays in commemoration of N. W. English, (Athlone, Old Athlone Society: 1980), p.

118; Loeber, R., The Geography and Practice of English Colonisation in Ireland from 1534 to 1609,

(The Group for Study of Irish Historic Settlement: 1991), p. 19 & 34-36, the latter of which again gives

passing mention and, although inaccurate, description of Dundonnell;& O‟Conor, K., “English settlement

and change in Roscommon during the late sixteenth and seventeenth centuries”, in Horning, A., Ó Baoill,

R., Donnelly, C., Logue, P.,(eds.) The Post Medieval Archaeology of Ireland 1550-1850, (2007), pp.

189-203, p. 192 213

Moore, Dr. A., “Mac Keoghs of Moyfin”, Old Athlone Society Journal, No. 4 (1974), pp. 56-70, p. 60 214

Archaeological Survey of Ireland, 2010, National Monuments Service, Ref. No. RO051-047001

(Description) 215

Barry, The Archaeology of Medieval Ireland, p. 186

93

7.3 - The outer defences at Dundonnell

There are no visible surface remnants of any stonework on the banks defining the

enclosure at Dundonnell. This strongly suggests that the perimeter defences of the site

during the stronghouse phase were built of timber, possibly accompanied by a timber

gatehouse. Croinin has suggested that a number of seemingly reused ringforts were

occupied by English settlers post 1572, and were subsequently burned during the Nine

Years War.216

This increases the possibility that earth and timber bawns were still very

much in use to defend sites at this late date in Roscommon. This practice wasn‟t

confined to Roscommon either, as numerous examples from across the country for

fortified and stronghouses show perimeter defences constructed of earth and timber,

even cob or thick quickset thorn hedges in some instances.217

Therefore, be it a 16th

-

century English colonial site, or a 17th

-century defended residence of the Mac

Keogh218

, Dundonnell shows that not all defended masonry sites in this period had

masonry defences on the perimeter.

7.4 - Stronghouses in Ireland

The first building types to be developed from the earlier tower house in the Irish rural

landscape was that of the fortified house and stronghouse, both of which began to be

constructed in roughly in the mid-to-late 16th

and early-to-mid 17th

-centuries. A

discussion of tower houses is too broad to be making general remarks on here, save to

216

Cronin, “The Elizabethan colony in Co. Roscommon”, p. 118 217

Hill, G., An historical account of the plantation of Ulster at the commencement of the seventeenth

century, 1608-1620 (Belfast: 1877; reprint, Shannon: 1970), pgs. 481-561; Nicholls, K., “Gaelic society

and economy”, Cosgrove, A., (ed.) A New History of Ireland II: Medieval Ireland 1169-1534, (Oxford:

2008), p. 405; Cairns, C. T., Irish Tower Houses: a County Tipperary case study, (Athlone: 1987), p. 17 218

Moore, “Mac Keoghs of Moyfin” [Continued], p. 60

94

note Sweetman‟s 1999 book The Medieval Castles of Ireland for its informative study

of the tower house.219

The tower house provided a step away from the heavily

defended castles that dominated key positions in the landscape, allowing lesser men of

wealth to show their status by constructing a form of castellated house on their lands.

They were essentially a compromise between the aforementioned castle and the

undefended Elizabethan mansions that signalled the end of the castle as a valid

building type.

A further progression away from outright defensibility comes with the appearance of

the fortified house in the 16th

and early 17th

-centuries, where definite aspects of

architectural design and style become the order of the day. Aesthetically pleasing, with

increasingly comfortable living quarters, while retaining some features that imply that

it was designed with a certain degree of defence, without it being the primary concern.

Sweetman describes the fortified house as:

“usually symmetrical, contained large mullioned and transomed windows, had high

gables and massive lozenge or diamond-shaped chimney stacks.”220

Examples of fortified houses predominate in the more settled lands of Leinster and

Munster, with fine examples including Rathfarnham Castle, Burncourt, Co. Tipperary,

and Kanturk, Co. Cork. Often tower houses had fortified houses constructed onto the

original building, and the tower house was often incorporated into the Renaissance-

based designs of the fortified house, such as at Athlumney, Co. Meath, and Donegal

219

Sweetman, The Medieval Castles of Ireland, pgs. 137-174 220

Sweetman, The Medieval Castles of Ireland, p. 175

95

Castle, built in the early 17th

-century. For Connacht, we have a few interesting

examples of fortified houses that vary in terms of refinement, ranging from Portumna

Castle, Co. Galway to the fortified house with attached tower house at Deel, Co.

Mayo.221

There are also some examples extant from Roscommon, specifically in the

south, with Gort and Athleague some of the most prominent.222

If we shift our view from the fortified house to the stronghouse, therefore, we get

picture that on the surface seems to be just a less impressive version of the fortified

house. Sweetman describes the stronghouse as:

“far less impressive than the fortified houses and are usually only two storeys high

with the ground floor defended by slit opes. The first floor is slightly more

commodious, having a fireplace and larger plain windows. The chimneystacks are not

impressive and are set on either gable. Few of the stronghouses retain their bawn walls

but presumably they all had some type of outer defensive works. They also lack other

defensive features such as bartizans and machicolations, crennellations and mural

stairs and passages. They are in fact, as their name describes them, „stronghouses‟.

They mark the end of castle building in Ireland.”223

Sweetman‟s definition is one of the most extensive available on the subject of

stronghouses; however, I feel it is fundamentally flawed. The first sentence provides

the context from which he based his definition. The contrast was made instantaneously

221

All above examples from Sweetman, The Medieval Castles of Ireland, pgs. 175-193 222

O‟Conor, “English settlement and change in Roscommon during the late sixteenth and seventeenth

centuries”, p. 191 223

Sweetman, The Medieval Castles of Ireland, pgs. 175-176

96

between the Renaissance-influenced fortified house and the „far less impressive‟

stronghouse, and from there it seems to have merely been an exercise in what was less

remarkable about the stronghouse. I think that if the stronghouse and fortified house

are viewed separately in terms of design and rather discussed in terms of function and

role, we may have a more rounded look as to why the stronghouse was to feature so

heavily on the landscape of the midlands224

in the 16th

and 17th

-centuries. Sweetman

correctly asserts that while stronghouses are roughly contemporary with fortified

houses, they are in fact mainly confined to the first quarter of the 17th

-century, if not

the late 16th

-century, while, in some cases at least, the fortified house has a broader

and later timeline.

The theory I would like to put forward is based on the fact that in the last decades of

the 16th

-century, with the hostile nature of the midlands meaning that a fortified house

might not necessarily be the most prudent option as regards defence, the somewhat

more militarily and practically orientated stronghouse was indeed a better option. The

possibility also exists that this stronghouse was reoccupied in the early 17th

-century by

the native Mac Keogh landowners225

, but it wasn‟t deemed important enough at this

time to record on the Stafford Survey map for the Barony of Athlone in 1636. (Fig. 4)

7.5 - Other stronghouses in Roscommon

Along with Dundonnell, the masonry buildings at Cloonbigney, the recorded but

destroyed Tully, and Lowberry all fall into the category of stronghouse, according to

224

Sweetman, The Medieval Castles of Ireland, pgs. 193-194 225

Moore, Dr. A., “Mac Keoghs of Moyfin”[Continued], p. 60

97

Sweetman.226

With the cases of Tully and Cloonbigney both being in close proximity

to Dundonnell, it may indicate something of society in the period, as, at least in the

case of Cloonbigney, we know that the house was constructed by one of Malby‟s

Ulster cavalrymen, John Moore, on the site of a native tower house, to protect his

newly acquired lands in the years leading up to the Nine Years War.227

7.6 - Conclusions

Therefore to conclude this chapter, all that is left to be done is to attempt to provide a

new definition for the castle type that is the stronghouse, mindful to keep it separate

from any definition of the fortified house.

The stronghouse is one of the new, and final classes of masonry castle building in

Ireland and centres around the mid-16th

to mid-17th

-century. It is typically a semi-

defended house, often with either a bawn of timber or stone. It is often built with a

balance of gun loops and more architecturally influenced windows, but with the

ground floor well serviced defensively. It is usually of two-storey construction and

built in a practical rather than a commodious fashion, potentially portraying something

of its use. With stout walls and simple, usually timber partition interior, the

stronghouse is definitely built less for comfortable living than acting as a class of

defensive outpost of sorts. The military character of these buildings along with their

location in the hostile midlands stands at odds with the fortified houses that proliferate

226

Sweetman, The Medieval Castles of Ireland, pgs. 197-198 227

Moore, Dr. A., “Mac Keoghs of Moyfin”[Continued], p. 60 & Cronin, “The Elizabethan colony in Co.

Roscommon”, p. 115

98

in other parts of the country at this time, the stronghouse truly seen as the end point of

castle building in Ireland.

The presence of the stronghouse in the landscape shows a clear need for gentry to have

to defend themselves at such a late date. Indeed the existence of the hall house, a

building type not dissimilar to the stronghouse in design, are a reminder that minor

gentry, (who would have lived in undefended manor houses over most of

contemporary England) needed this protection from attack from the beginnings of the

colonial enterprise in Ireland during the late 12th

-century, right down to the mid 17th

-

century.

99

Chapter 8- Conclusions

8.0 - Introduction

The conclusion of this thesis serves to show whether the aims of the thesis have been

completed successfully, to highlight any future research that needs to be undertaken to

further the understanding of Dundonnell Castle and castle studies in general, in both

the archaeological and historical disciplines.

8.1 - Main conclusions of the thesis

The primary aim of this thesis was to gain a better understanding of Dundonnell and

its functions, along with how it developed through time. By approaching the site with

regard to its three potential phases, with both an historical and archaeological approach

in mind, I believe Dundonnell received the study it was due. The postulations of both

ringfort and ringwork phases were based upon solid archaeological foundations, with

the addition of historical and siting evidence serving to show how viable the

aforementioned phases would have been to the occupiers of the site in both periods.

The possibility of it being recognised by Tudor colonisers as a Crown castle and thus

refortified and reoccupied seems likely, based on parallel practice in the surrounding

and wider area for the same period. As a result, we now have a much more detailed

account of the often overlooked medieval history of the area, something that studying

Dundonnell has gone a small way to rectifying.

Another aim of the thesis was to further the study of the castle type that is the

stronghouse. This was achieved through the detailed plans and analysis of the

building, along with comparisons with similar sites, thus clarifying its function as

100

separate from that of the fortified house. I believe we now have a much more concrete

knowledge of the stronghouse and hopefully some idea as to why it may have been

built, when analysing it alongside the histories for the period of its occupation.

Which brings us to the plans and descriptions of the site itself. This endeavour was

deemed important, as it had never been done before with regard to Dundonnell,

something that is all too frequently seen with Irish castle studies. Therefore the most

up-to-date information-gathering equipment was employed to provide the earthwork at

Dundonnell with a modern plan, while the description and plan of the stronghouse was

achieved by more traditional, but equally as scientific, methods. Combined they give

us an understanding of the site that otherwise would be sadly obscured.

8.2 - Future work

More work needs to be done on the stronghouse as a castle type. The architectural and

functional links that the stronghouse has to the bastle houses of the Scottish Borders is

something that, due to space constraints, could not be pursued for this study, however,

is something that should be explored more.

Of course, excavation of the site must be undertaken in order to prove the theories

outlined and argued in this thesis conclusively.

Finally, I felt that this thesis was a small step in the right direction of pursuing a

multidisciplinary approach in studying the past, something that has to be built upon

substantially, in order to get the clearest and most detailed picture of our past,

irregardless of discipline.

101

Primary Source Bibliography Annála Connacht: the Annals of Connacht, AD. 1224-1544, Freeman, A.D., (ed.),

(Dublin: 1944; reprint 1977)

Calendar of documents relating to Ireland, 1252-1284, (vol. 2), Sweetman, H. S.

(ed.), (Nendeln, 1974)

Calendar of State Papers relating to Ireland of the reigns of Henry VIII, Edward

VI, and Elizabeth, 1509-[1603]. Hamilton, H. C. (ed.). 11 vols. P.R.O. (London:

1920-31)

Cromwellian Grantees 1654-1658: County Roscommon Tagh Mc : Connell Parish

Athlone Barony, recorded in Taughmaconnell: a history, (Athlone: 2000)

Gerald of Wales, The History and Topography of Ireland, trans. O‟Meara, J. J.,

(Penguin, London: 1982)

The Annals of Clonmacnoise: Being annals of Ireland from the earliest period to

A.D. 1408, Murphy, Rev. D., (ed.), (Dublin: 1896, facsimile reprint 1993)

Annals of the kingdom of Ireland by the Four Masters: from the earliest period to

1171, O‟Donovan, J., (ed.), (Hodges and Smith, Dublin: 1854)

102

Secondary Source Bibliography

Archaeological Survey of Ireland, 2010, National Monuments Service [online],

available: http://webgis.archaeology.ie/NationalMonuments/flexviewer/

[accessed 9th

June 2011]

Barrett, G. F., Graham, B. J., Lynn, C. J., “Some Considerations concerning the

Dating and Distribution of Ring-Forts in Ireland”, Ulster Journal of

Archaeology, Third Series, Vol. 38 (1975), pp. 33-47

Barry, T., The Archaeology of Medieval Ireland, (London & New York: 1987)

Byrne, F. J., Irish Kings and High-Kings, (Dublin: 2001)

Byrne, F. J., “The trembling sod: Ireland in 1169”, in Cosgrove, A., (ed.) A New

History of Ireland II: Medieval Ireland 1169-1534, (Oxford: 2008)

Cahill, T., How the Irish Saved Civilisation, (London: 1995)

Cairns, C. T., Irish Tower Houses: a County Tipperary case study, (Athlone:

1987)

Collins, T., & Cummins, A., “Excavation of a Medieval Ringwork at

Ballysimon County Limerick”, Aegis Archaeology Reports 1 (2001)

Comber, M., “Trade and Communication in Early Historic Ireland”, The

Journal of Irish Archaeology, Vol. 10 (2001), pp. 73-92

Cosgrove, A., “Ireland beyond the Pale, 1399-1460”, in Cosgrove, A., (ed.) A

New History of Ireland II: Medieval Ireland 1169-1534, (Oxford: 2008)

Craig, M., The Architecture of Ireland from the earliest times to 1880, (London

& Dublin: 1989)

Creighton, O. H., Castles and Landscapes: Power, Community and

Fortification in Medieval England, (2002)

Cronin, T., “The Elizabethan colony in Co. Roscommon”, pp. 107-20, in

Murtagh, H. (ed.), Irish midlands studies-essays in commemoration of N. W.

English, (Old Athlone Society, Athlone: 1980)

De Meulemeester, J., O‟Conor, K., “Chapter 11: Fortifications”, in The

Archaeology of Medieval Europe: Vol. – Eighth to Twelfth Centuries AD,

(Aarhus: 2007), pp. 316-341

Dolley, M., Anglo-Norman Ireland (The Gill History of Ireland 3: 1972)

Doran, L., “Medieval Communication Routes through Longford and

Roscommon and Their Associated Settlements”, Proceedings of the Royal Irish

Academy. Section C: Archaeology, Celtic Studies, History, Linguistics,

Literature, Vol. 104C, No. 3 (2004), pp. 57-80

Dudley Edwards, R., “Anglo-Norman Relations with Connacht, 1169-1224”,

Irish Historical Studies, Vol. 1, No. 2 (Sep., 1938), pp. 135-153

Duffy, P. J., Edwards, D., & FitzPatrick, E., Gaelic Ireland c. 1250-1650: Land,

Lordship and Settlement, (Four Courts Press, Dublin: 2001)

Edwards, N., The Archaeology of Early Medieval Ireland, (London: 1990)

Egan, P. K., “The Carmelite Cell of Bealaneny”, Journal of the Galway

Archaeology and Historical Society, Vol. 26, No. ½ (1954/1955)

Finan, T., (ed.), Medieval Lough Ce – History, Archaeology and Landscape,

(Dublin: 2010)

103

Graham, B., “Medieval Timber and Earthwork Fortifications in Western

Ireland”, Medieval Archaeology (1988b), pp. 110-129

Graham, B., “Medieval Settlement in County Roscommon”, Proceedings of the

Royal Irish Academy. Section C: Archaeology, Celtic Studies, History,

Linguistics, Literature, Vol. 88C (1988), pp. 19-38

Herity, M., “Motes and Mounds at Royal Sites in Ireland”, The Journal of the

Royal Society of Antiquaries of Ireland, Vol. 123 (1993), pp. 127-151

Higham, R. A., “Early Castles in Devon, 1068-1201”, Château Gaillard 9-10,

(1982), pp. 102-115

Higham, R., Barker, P., Timber Castles, (London: 1992)

Hill, G., An historical account of the plantation of Ulster at the commencement

of the seventeenth century, 1608-1620, (Belfast: 1877; reprint, Shannon: 1970)

Hinton, D. A., Archaeology, Economy and Society: England from the Fifth to

the Fifteenth Century, (London: 1990)

Jope, E. M., Jope, H. M., & Johnson, E. A., The Archaeological Survey of Co.

Down, (Belfast: 1966)

Kelleher, J. V., “Uí Maine in the annals and genealogies to 1225”, Celtica, ix

(1971), pp. 61-112

Kenyon, J. R., Medieval Fortifications, (Leicester: 1996)

King, D. & Alcock, L., “The Ringworks of England and Wales” in Château

Gaillard 3, (1969), pp. 90-127

Leask, H. G., Irish Castles and Castellated Houses, (Dundalk: 1941)

Liddiard, R., “Landscapes of Lordship: Norman Castles and the Countryside in

Medieval Norfolk, 1066-1200”, BAR British Series 309 (2000)

Loeber, R., The Geography and Practice of English Colonisation in Ireland

from 1534 to 1609, (The Group for Study of Irish Historic Settlement: 1991)

Lydon, J., “A land of war”, in Cosgrove, A., (ed.) A New History of Ireland II:

Medieval Ireland 1169-1534, (Oxford: 2008)

MacCotter, P., Medieval Ireland: Territorial, Economic and Political Divisions,

(Dublin: 2008)

Mac Shamhráin, A.S., “Placenames as Indicators of Settlement”, Archaeology

Ireland, Vol. 5, No. 3 (Autumn 1991), pp. 19-21

Mallory, J. & McNeill, T., The archaeology of Ulster, (Belfast: 1991)

Martin, F. X., “John, Lord of Ireland, 1185-1216”, in Cosgrove, A., (ed.) A New

History of Ireland II: Medieval Ireland 1169-1534, (Oxford: 2008)

McAffe, P., Irish Stone Walls, (Dublin: 1997)

McNeill, T. E., Castles in Ireland: Feudal Power in a Gaelic World, (London

and New York: 1997)

Morrissey, J., “Cultural geographies of the contact zone: Gaels, Galls and

overlapping territories in late Medieval Ireland”, Social and Cultural

Geography, Vol. 6, No. 4, August 2005, pp. 551-566

Moore, Dr. A., “The Cell of the Canonesses Regular of St. Augustine at

Clonoghill”, Old Athlone Society Journal, No. 1,(1969), pp. 15-16

Moore, Dr. A., “Mac Keoghs of Moyfin”, Old Athlone Society Journal, Vol. I,

No. 4, (1974-75), pp. 234-237

104

Moore, Dr. A., “Mac Keoghs of Moyfin”, [Continued], Old Athlone Society

Journal, Vol. II, No. 5, (1978), pp. 56-70

Moore, M., (ed.), Taughmaconnell: a history, (Athlone: 2000)

Murphy, M., “Roscommon Castle: Underestimated in Terms of Location?”,

Journal of the Galway Archaeological and Historical Society, Vol. 55 (2003),

pp. 38-49

Murphy, M. & O‟Conor, K., Roscommon Castle: A Visitor’s Guide,

(Roscommon County Council: 2008)

Nicholls, K., Gaelic and Gaelicised Ireland in the Middle Ages, (The Gill

History of Ireland 4: 1972)

Nicholls, K., “Gaelic society and economy”, Cosgrove, A., (ed.) A New History

of Ireland II: Medieval Ireland 1169-1534, (Oxford: 2008)

O‟Conor Don, C.O., The O’Conors of Connaught, (Hodges & Figgis, Dublin:

1891)

O‟Conor, K., The Archaeology of Medieval Rural Settlement in Ireland,

(Dublin: 1998)

O‟Conor, K., “English settlement and change in Roscommon during the late

sixteenth and seventeenth centuries”, in Horning, A., Ó Baoill, R., Donnelly, C.,

Logue, P., (eds.) The Post Medieval Archaeology of Ireland 1550-1850, (2007),

pp. 189-203

O‟Conor, K., “Castle Studies in Ireland”, Chateau Gaillard 23 (2008), pp. 329-

39

Ó Corráin, D., Ireland before the Normans (The Gill History of Ireland 2: 1972)

Ó Corráin, D., “Ireland c.800: aspects of society”, in Ó Cróinín, D., (ed.), The

New History of Ireland 1: Prehistoric and Early Ireland, (Oxford: 2008)

Ó Cróinín, D., “Ireland 400-800”, in Ó Cróinín, D., (ed.), The New History of

Ireland 1: Prehistoric and Early Ireland, (Oxford: 2008), pp. 182-234

O‟Donovan, J., The Tribes and Customs of Hy-Many, commonly called

O’Kelly’s country, (Dublin: 1843)

O‟Donovan, J., “The Registry of Clonmacnoise: With Notes and Introductory

Remarks”, The Journal of the Kilkenny and South-East of Ireland

Archaeological Society, New Series, Vol. 1, No. 2 (1857), pp.444-460

Ó Drisceoil, C., “Recycled Ringforts: The Evidence from Archaeological

Excavation for the Conversion of Pre-Existing Monuments to Motte Castles in

Medieval Ireland”, Journal of the County Louth Archaeological and Historical

Society, Vol. 25, No. 2 (2002), pp. 189-201

O‟Faolain, S., The Great O’Neill, (Mercier Press, Cork 1997)

O‟Keefe, T., “The Fortifications of Western Ireland, AD 1100-1300, and Their

Interpretation”, Journal of the Galway Archaeological and Historical Society,

Vol. 50 (1998), pp. 184-200

Ó Lochlainn, C., “Roadways in Ancient Ireland”, in Ryan, J., (ed.)

Féilsgribhinn Eóin Mhic Néill, (Dublin: 1940), pp. 465-474

Ó Ríordáin, S., Antiquities of the Irish countryside, 5th

ed. (London: 1979)

Ordnance Survey Discovery Series, 2nd

Edition, 47

Orpen, G. H., “Athlone Castle: Its Early History, with Notes on some

Neighbouring Castles”, The Journal of the Royal Society of Antiquaries of

105

Ireland, Fifth Series, Vol. 37, No. 3, [Fifth Series, Vol. 17] (Sep. 30, 1907), pp.

257-276

Orpen, G. H., “The mote of Lisardowlan, County Longford”, The Journal of the

Royal Society of Antiquaries of Ireland, Fifth Series, Vol. 40, No. 3, [Fifth

Series, Vol. 20], (Sep. 30, 1910), pp. 223-225

Orpen, G. H., Ireland under the Normans, (Dublin: 2005)

Prunty, J., Maps and Map Making in Local History, (Dublin: 2004)

Quinn, D. B., “ „Irish‟ Ireland and „English‟ Ireland”, in Cosgrove, A., (ed.) A

New History of Ireland II: Medieval Ireland 1169-1534, (Oxford: 2008)

Quinn, D. B., “The reemergence of English policy as a major factor in Irish

affairs, 1520-34”, in Cosgrove, A., (ed.) A New History of Ireland II: Medieval

Ireland 1169-1534, (Oxford: 2008)

Spurgeon, C.J., “The castles of Glamorgan: some sites and theories of general

interest”, Château Gaillard 13, (1978b), pp. 203-226

Stout, M., The Irish Ringfort, (Four Courts Press: Dublin: 1997)

Sweetman, D., Irish Castles and Fortified Houses, (Dublin: 1995)

Sweetman, D., The Medieval Castles of Ireland, (Cork: 1999)

The Story of Ireland (2011) RTÉ, March. 10.15pm

Verstaten, F., “Both King and Vassal: Feidlim Ua Conchobair of Connacht,

1230-65”, Journal of the Galway Archaeological and Historical Society, Vol.

55 (2003), pp. 13-37

Walsh, P., “Christian Kings of Connacht”, Journal of the Galway

Archaeological and Historical Society, Vol. 17, No. ¾ (1937), pp. 124-143

Walsh, P., “Connacht in the Book of Rights”, Journal of the Galway

Archaeological and Historical Society, Vol. 19, No. ½ (1940), pp. 1-15

Walton, H., “The English in Connacht, 1171-1333”, (Unpublished PhD Trinity

College Dublin: 1980)

Warner, P., The Medieval Castle: Life in a Fortress in Peace and War, (Penguin

Classic History: 2001)

Watt, J. A., The Church and the Two Nations in Medieval Ireland, (Cambridge:

1970)

Watt, J. A., “The Anglo-Irish colony under strain, 1327-99”, in Cosgrove, A.,

(ed.) A New History of Ireland II: Medieval Ireland 1169-1534, (Oxford: 2008)

Weld, I., Statistical Survey of the County of Roscommon, drawn up under the

directions of the Royal Dublin Society, (Dublin: 1832)