2018

MULTI-METHOD REMOTE SENSING SURVEY AT LOUGH CROAN, CO. ROSCOMMON – RCC HERITAGE RESEARCH BURSARY 2018 Project Report

DANIEL CURLEY, 31ST OCTOBER 2018 Table of Contents Table of Figures ...... 2 Introduction ...... 3 Lough Croan – Physical Description...... 4 Placename Evidence ...... 4 Communication routes ...... 7 Historical background ...... 8 Artefactual assemblages from Lough Croan...... 10 Focussed archaeological research at Lough Croan ...... 12 Discussion of the Preliminary Results of the Remote Sensing Investigations ...... 19 Future Work ...... 22 Acknowledgements ...... 23 Bibliography ...... 24

1

Table of Figures Figure 1 - Contoured topographical plan of Lough Croan, indicating recorded archaeological monuments, and principal placenames. (LiDAR data courtesy of OSi) ...... 7 Figure 2 - Digital modelling of the water level of Lough Croan turlough to 69m OD. Note the range of 'islands' located on the lake, particularly the collection of landmarks at the centre of the lake. The red arrow indicates the location targeted for remote sensing investigation. (Data courtesy of OSi, image generated by LGS) ...... 12 Figure 3 - Location of area chosen for investigation, outlined in red on a six-inch OSi map with plotted location of RMP RO044-089--- to the west. (Map OSi/ASI) ...... 13 Figure 4 - LiDAR image of the survey area, marked by a rectangle (LiDAR data courtesy of OSi, acquired courtesy of RIA); Figure 5 - Location of area considered for investigation, with mound outlined in red on an aerial photograph, with plotted location of RMP RO044-089--- to the west. (Photo: Digital Globe)...... 14 Figure 6 - Contoured topographical plan of the target location, derived from LiDAR data (Data courtesy of OSi, image generated by LGS) ...... 14 Figure 7 - Earth Resistance Survey in progress at Lough Croan, with Total Station in foreground. (Image courtesy of Kevin Barton) ...... 15 Figure 8 - Reconnaissance survey geo-referencing the groundfast stones on the summit of the promontory at Lough Croan, set within the survey grids. (LiDAR data courtesy of OSi, image generated by LGS) ...... 16 Figure 9 - Electrical Resistivity Tomography (ERT) survey conducted over the summit of the promontory at Lough Croan. (Image courtesy of LGS) ...... 17 Figure 10 - Earth Resistance Survey conducted in 10m x 10m grids. (Image courtesy of LGS) ...... 18 Figure 11 - Magnetic Susceptibility Survey conducted in 10m x 10m grids. (Image courtesy of LGS) . 19

2

Introduction This report is being submitted to the Heritage Office, towards the partial fulfilment of the conditions of my grant of a Heritage Research Bursary for 2018. The report begins with a detailed historical background for the Lough Croan landscape, which provides justification for the targeting of Lough Croan for the application of the multi-method remote sensing survey in the landscape.

This is followed up by a consideration of the recently acquired LiDAR dataset of a substantial area of the Lough Croan landscape, which the author received thanks to a Royal Irish Academy grant award. This LiDAR dataset has enabled the author to produce a Digital Terrain Model (DTM) of the Lough Croan landscape, which has provided a detailed topographical basemap from which to study the archaeological monuments at Lough Croan. This has uncovered a number of anomalies which may prove to be unrecorded monuments, while also enabling the author to digitally ‘flood’ the turlough at Lough Croan, in order to better understand the archaeological remains within the former lake, as well as its near vicinity. This data has been vital in terms of identifying a specific monument for multi-method remote sensing survey, and provided the evidence required to target the monument that was studied thanks to the funds of the research bursary.

The survey will then be discussed, and some preliminary results and discussion will be communicated. These results have not been fully processed to date, therefore the conclusions provided in this report may not be the conclusions arrived at upon the final submission of the PhD research.

As mentioned, this research is part of this author’s PhD research, which is being carried out in NUI Galway, and is based on an exploration of the archaeology of the Úa Cellaig (O’Kelly) lordship of Uí Maine, c. 1100-1600. The research focusses on the understudied later medieval O'Kelly subkingdom and later lordship of Uí Maine. At its height in the fifteenth century, this lordship consisted of large parts of east Galway and south Roscommon. The principal aim of the research, taking a multi- disciplinary approach using the available archaeological, remote sensing, historical and literary evidence, is really to chart and trace the formation, character, physical appearance and development of Uí Maine from the twelfth century through to the demise of the O’Kelly lordship in the late sixteenth and early seventeenth centuries. There are two key aims of the author’s research which the bursary will aid substantially in achieving. One is to try and identify and investigate the evolution of the settlement archaeology of the O’Kelly lordship through time - from ringfort, crannóg and cashel, to towerhouse and other fortified residences of the late medieval/early modern period. The second is to identify the cenn áiteanna or lordship centres of the O’Kelly kings and lords of Uí Maine in the high and late medieval period. A wealth of hitherto-untapped archaeological (including unpublished and partially published excavation reports linked to infrastructural projects), historical and literary sources exists for Uí Maine and its adjacent areas. The available rich cartographic evidence (such as the Strafford map of the 1630s), combined with much of the data from the latter sources, would suggest that the landscape of later medieval Uí Maine can be reconstructed.

The available source material will then be supplemented by desk and fieldwork-based methodologies, in order to develop as complete a picture of later medieval Uí Maine as possible, which will then offer a range of opportunities for development in the area. This author intends on using the research as a key initiator in the development of rural tourism initiatives in the south Roscommon/east Galway region, one that has been identified as deficient in terms of harnessing its tourism potential. Through my experience of working in a similar area of high

3 rural tourism potential, I can identify outputs that can serve as templates for similar projects in the research area, with themes including but not exclusive to archaeological, historical and heritage tourism, outdoor and walking tourism and genealogy. Lough Croan – Physical Description The turlough of Lough Croan has an approximate area of 106.9ha (Goodwillie, 1992, p. 148), of which 106.2 ha lies within the boundary of the Lough Croan Turlough Special Area of Conservation (SAC). The turlough measures approximately 3.4 km at its furthest extents, and 570 m at its widest point. The elevation in the basin of the turlough is approximately 67.5 m OD, while the elevation of the surrounding glacial hills approaches 104 m to the south and 106 m to the north of the turlough, located in the townlands of Cronin and Correal respectively.

The underlying geology of the Lough Croan area, and indeed Co. Roscommon as a whole, is primarily one of Carboniferous Limestones (Parkes, Meehan, & Préteseille, October 2012, p. 18), while the soil composition of the turlough bed is primarily peaty, with silt and an impure marl in places (Goodwillie, 1992, p. 149). The chemical composition of soils in most of these limestone regions contains an abundance in soil nutrients which are very complementary to high-quality cattle production (Collins, 2016, p. 2). Most of the Lough Croan area is open to grazing, particularly of sheep, but the reed beds of the turlough are generally protected by field fences and boundary banks (Goodwillie, 1992, p. 149). Around the margins of the turlough, grassland is dominant, with calcareous grassland at the southern end where the shore rises sharply, and elsewhere the grassland is more nutrient-enriched and less species diverse (DAHG, 2013). Placename Evidence The townlands which include Lough Croan turlough in their limits, in a clockwise selection from the northwest, include: Carrowntlieve – Ceathrú an tSléibhe (the quarterland of the mountain), Lisnagavragh – Lios na gCabhrach (the ringfort of the embossed shields), Grange – An Ghráinseach (the grange or monastic farm), Ballylion – Baile Uí Laighin (the town of Ó Laighin), Coolegary – Cúl a gharaidh (back of the garden), Cornalee – Corr na Lao (the round hill of the calves), Garrynphort [two townlands] – Garraí an Phoirt (the garden/court of the landing-place), Turrock – An Turrac [originally Caisleán an Turraic] (the castle of the attack), Liswilliam – Lios Uilliam (William's fort), and Carrownurlar – Ceathramhadh an urláir (quarter of the floor), also recorded as Ceathramadh caol (narrow quarter) (DCHG, 2008-2018).

Beyond the limits of the turlough itself, some of the townlands in the immediate environs preserve some intriguing placenames that may provide information on the activity of past societies at Lough Croan. Thematically, and where the townland names are not describing physical features of the landscape, the names are concerned with martial associations, assembly, ownership, folkloric devices, and cattle production.

The martial associations are most readily seen with the aforementioned Lisnagavragh and Turrock. Assembly or periodic fairs are recorded with the placename Tullyneeny – Tulaigh an aonaigh (hill of the cattle fair). Tullyneeny is located 2.5 km due south of the turlough, on a prominent rise on a NNW-SSE esker ridge, and is bounded to the north by the townland of Gortaphuill – Gort a phoill (field of the hole). Within Gortaphuill townland there is a landmark known as ‘Fair Hill’, c. 73 m OD, and which is topped today by a wayside cross (RO047-026----). The First Edition Ordnance Survey maps locates a standing stone at this location, and ‘Fair Hill’ is located c. 212 m west of a moated site (RO047-028----), in the adjacent townland of Cuilleenirwan – Coillín Íriomháin (Irvine's little wood). This collection of features is worthy of further investigation, in order to ascertain the links between

4 the placename, the low hill of ‘Fair Hill’, which was once topped with a standing stone in Gortaphuill townland, and the presumably high medieval moated site in Cuilleenirwan.

The townland names of Ballylion and Liswilliam record a surname and personal name respectively, and the latter placename may refer to the substantial oval-shaped ringfort (RO044-065001-) located in the North West quadrant of the townland. Liswilliam townland is located to the west of Lough Croan, and the townland of Turrock, which is the historically-attested location for an O’Kelly castle in 1536. The aforementioned ringfort is the only such monument in the townland, and the townland name could potentially refer to one of only two O’Kelly kings of Uí Maine (Uilliam Buide, king of Uí Maine 1349-1381; or William, king of Iar Uí Maine 1472–1476, all Uí Maine 1476–1487). If the placename refers to either, then it is most likely to have referred to Uilliam Buide, who perhaps constructed or modified the ringfort at some point during his fourteenth century reign. The Anglo- Norman personal name William seems to be first adopted by the O’Kelly dynasts with the aforementioned Uilliam Buide.

Folklore and mythology is represented in the environs of Lough Croan in the form of the townland names of Carricknagat – Carraic na gcat (rock of the [wild] cat) and Lissaphuca – Lios a' phúca (pook's or goblin's fort), located to the north west and south west of Lough Croan respectively. It is even possible that the townland name Carricknagat may preserve links to an archaic tribal grouping known as the Cattraige. As Gleeson observes:

‘This obscure polity had a number of branches settled in different districts throughout early medieval but commonly associated with royal landscapes, and would appear to have been ritual specialists who discharged religious functions, including the inauguration of kings’ (Gleeson, 2018, pp. 189-190).

The Cattraige are identified as being one of the dáerthuatha or ‘tributary peoples’ of the Uí Maine, and held lands either side of the River Suck (Ó Riain, 1997, p. 701), and these Cattraige are posited by Gleeson as potential ritual specialists, a role which they may have carried from pre-Christian times into the historic period (Gleeson, 2018, p. 191). The record for a megalithic structure (RO044- 058----) within the townland limits, coupled with the proximity of the townland, and indeed the megalithic structure, to the adjacent townland of Carrowntemple - Ceathramhadh an teampuill (the church quarter), suggests as continuity of religious activity in this part of the Lough Croan landscape through time. This motif is also seen at Cashel, Co. Tipperary, and possibly Dunseverick, Co. Antrim (Gleeson, 2018, pp. 190-192).

The last group of townland names which this author wishes to address relates particularly to names associated with cattle, a trend which may provide an important insight into the focus of economic activity in the Lough Croan environs during the medieval period. The list begins with the aforementioned names of Tullyneeny and Cornalee, both located to the south of Lough Croan turlough. To this can be added the townland name of Lisseenamanragh – Lisín na mannrach (little fort of the stalls or mangers), a name which corresponds directly with the archaeological remains (RO047-035001-; RO047-035002-; RO047-036----; RO047-037----; RO047-038----), a combination of stone enclosures and field systems, designed to house livestock. Lisseenamanragh townland is 3.5 km away from Lough Croan turlough as the crow flies.

This author would like to speculatively include a further townland name to this list, the townland of Grange. Grange bounds onto the turlough of Lough Croan, near its south-eastern end, to the north of the turlough limits, and is 4.06 km² in area. This author believes that the O’Kelly manuscript of Leabhar Ua Maine may contain a reference to the placename in the elongated and now extinct form

5 of Gráinseach Chairn Bhuaileadh (Ó Muraíle, 2010, p. 64). This placename translates to the ‘Grange of the cairn of the cattle-enclosure’, (Nollaig Ó Muraíle pers. comm.). Grange (An Gráinseach) contains five mounds of various classes, including a substantial bowl barrow (RO045-121001-), which may relate the use of the term cairn in the placename. Gráinseach Chairn Bhuaileadh is recorded as the place of death of a thirteenth century king of Uí Maine, Maine son of Conchobhair O’Kelly, who died in 1268 having reigned just half a year. Half a kilometre to the south of this barrow in the same townland is a large univallate ringfort, c. 54m in internal diameter (RO045-123----), which albeit speculative, may have been the residence and place of death being suggested as Gráinseach Chairn Bhuaileadh.

The grouping of townland names concerning cattle surrounding Lough Croan turlough is suggestive that this particular area was utilised extensively for the production and trading of livestock at some point in the past, and this trend seems to be repeated in another identified O’Kelly lordship cenn áit surrounding Galey Bay, Co. Roscommon. Here, the townland names of Cornamart – Cor na mart (round hill of the Beeves), Corboley – Corbuaile (odd booley or dairy, also possibly round hill of the dairy), Curraghalaher – Currach a' leathair (moor of the leather) and Pollalaher – Poll a leathair (hole of the leather), all bear evidence of a predominance in the practice of cattle rearing, butchery and processing of the associated resources (tanning and leather production) in the area.

Lough Croan turlough itself contains a number of placenames relating to the natural and manmade islands located on the former lake. All of the island names are preserved on the First Edition 6-Inch Ordnance Survey maps. Working from west to east, the first island name recorded is Edward’s Island. This is identifiable with the site of a crannóg (RO044-107----), one of five crannóga recorded on the Record of Monuments and Places (RMP) for the turlough. It is unclear who is being referred to in the island name, and this author’s only speculation is that it may relate to the father of a sixteenth century O’Kelly lord of Uí Maine, Éamonn O’Kelly, whose son Donnchadh was attacked at Turrock castle, presumably his residence, in 1536. Turrock House (RO044-114----), is a seventeenth- century house located on the site of the now ruined Turrock castle, presumably of form, situated less than 500 m from the crannóg, on what was once the shoreline of the lake. Irrespective of the connection between the placenames in this instance, there is a strong likelihood that Turrock castle and Edward’s Island were linked in terms of use, with the tower house operating as the dryland residence for the late medieval O’Kelly lords, with the crannóg acting as a refuge, guest accommodation, or location for feasting for the occupiers of the tower house, as well as a legitimiser of their antique origins in the area. Paul Logue has provided compelling evidence for the careful choices made by Aodh Ó Néill in terms of his places of residence, and the attributes Aodh sought to radiate to the various audiences, Gaelic and English, who were studying him in sixteenth century (Logue, 2018, pp. 269-292).

The next named island is Inchnaveague Island – Inse na bhFia (the island of the deer), which is also categorised as a crannóg (RO044-088----) The description attached to this monument records elements of a wooden protruding through the sod at the perimeter E-S-W. Located roughly at the centre of the turlough, Inchnaveague Island is in line of sight of a substantial univallate ringfort (RO044-072001-), c. 800 m to the north of the crannóg, perhaps the fort being described in the townland name in which it is located, Lisnagavragh. The name of the crannóg incorporates something of the activities that may have taken place on it. The reference to ‘deer’ implies that this crannóg site may have been used at some point in time for hosting guests and for feasting.

Located 200 m to the east of Inchnaveague Island is the named island of Illaunnamona –Oileán na Móna (The turf island). It is marked on the RMP (RO044-091----) but no longer deemed to be an archaeological site, instead a natural island, and is named for its physical appearance.

6

There are two further named islands located in the south-eastern end of the turlough, Inchnagower – Inse na nGabhar (Island of the Goats), and Inchnagreeve - Inse na gCraobh (Island of the Bush). While they are categorised as crannóga on the RMP, there are reservations over their identification as such. Local information states that bones were found at Inchnagower (RO044-094----), while there was a 7.9 m long dug-out canoe (RO044-124----) discovered and returned in situ between Inchnagower and Inchnagreeve (RO047-031----), which confirms that these two islands were in use at some point in the past.

Figure 1 - Contoured topographical plan of Lough Croan, indicating recorded archaeological monuments, and principal placenames. (LiDAR data courtesy of OSi) Communication routes Communication through the Lough Croan landscape was provided primarily by land routeways, due to its relative distance from the regional waterways of the River Suck, River Shannon and Lough Ree. Academic research into the premodern road systems of this part of the island is limited to two papers. The first, ‘Roadways in ancient Ireland’, was published in 1940 (Ó Lochlainn, 1940), and for an island-wide account of the routes of early medieval and, most likely, late , it is still unrivalled. More recently, a regional survey of the major and secondary routes of medieval Roscommon and Longford provides an in-depth analysis of the communication channels of this part of medieval Ireland (Doran, 2004), and Doran’s work highlights the presence of an important secondary route that connects the two major early medieval east-west routeways of the Slighe Assail and the Slighe Mhór. Named ‘Route 9’, this communication route leaves the Slighe Mhór or Eiscir Riada at Áth Nadsluaigh (the modern town of Ballinasloe), heading north, via Lough Croan, to Roscommon, and eventually Tulsk on the Slighe Assail (Doran, 2004, p. 72). From a logistical point of view then, the communities settling Lough Croan were very centrally placed in terms of accessing trade and transport routes both to the south, and to the north of them. This routeway was of vital

7 importance, considering as substantial areas to the west of Lough Croan and across the Suck into modern east Galway are served by treacherous and unnavigable peatland. To the east, the areas approaching Lough Ree, and south towards Athlone, possessed woodland and forest in abundance. Evidence of the difficulty that traversing this area caused can be seen with the 1273-74 reference to a pass being cut through the dense forest of the Feadha of Athlone in order to provide a vital communication route between the Anglo-Norman of Athlone and Rindoon (Anon, 1904, p. 41), a journey that previous would have presumably taken place by boat on Lough Ree. Historical background Lough Croan is present in a variety of historical and literary sources throughout the medieval period. Beginning in the eighth century, there is a possible very early allusion to Lough Croan recorded in the Annals of Tigernach and of the Four Masters. This occurs with the account of the battle of Belach Cró, adjudged to have taken place between the Clann Chremthainn of Uí Maine and the Delbna Nuadat, a battle won by the former group. O’Donovan suggests that the location described as Lochán-Bealaigh-Cró may be identifiable with Lough Croan (O'Donovan (ed. trans.), 1851-1856, pp. 354, note i).

The ninth-century Patrician hagiography Vita tripartita Sancti Patricii records ‘Loch Cróine’ very early in its Connacht sequence, and equates Lough Croan’s relocation to its present position with a miracle performed by St. Patrick (Stokes (ed. and trans.), 1887, pp. 85-87). It is interesting that Lough Croan is mentioned in this seemingly unrelated context.

The next possible reference to activity in the vicinity of Lough Croan occurs in the eleventh century. The year of 1074 is recorded in a series of chronicles for the death of Donnchad O’Kelly, king of Uí Maine, who was killed by his brother Tadg. The location of death is Inis Locha Caoláin, which Kelleher suggests may be a crannóg residence located on the double lakes of Cuilleenirwan Lough and Coolagarry Lough, just to the southeast of Lough Croan (Kelleher, 1971, p. 92). This identification remains to be confirmed.

The twelfth-century prose narrative Acallam na Senórach, the most important of the Fenian Cycle tales, initially seems to be an unexpected source of material on Uí Maine and Lough Croan. However, the geographical setting for the Acallam strongly advocates for a Roscommon locus for the composition (Connon, 2014, pp. 21-59), and an episode of the tale, the first meeting of Patrick with Muiredach Mór, King of Connacht, takes place in ‘the Land of Maine, and [to] Loch Linngáeth, now called Lough Croan’ (Dooley & Roe, 2008, p. 33). The ‘Land of Maine’ and Lough Croan are more than worthy of mention in the text, particularly as Patrick performs his most powerful miracle in this location, raising Áed, the king’s son, back to life following his collapse and death after a game of hurling (Dooley & Roe, 2008, p. 38). Muiredach Mór, and his son Áed, are regarded as literary representations of the high medieval king of Connacht, Cathal Crobhdearg Ua Conchobar and his son Áed respectively, and there is a body of evidence to suggest that Cathal Crobhdearg constructed a royal centre on the southern shore of Lough Croan, possibly in the northern end of Cornalee townland, at the site of an unclassified castle (RO041-001), known locally as an O’Kelly castle (Connon, 2014, p. 53).

If this castle, or another location in the area, is identifiable with a royal centre associated with Cathal Crobhdearg, it may be further evidence of the forced migration of the O’Kelly away from this part of Tír Maine at some point in the twelfth century. The aforementioned 1169 reference to the death of Ferchar Úa Fallamhain, chief of Clann Uadach and maor Ua Maine, from a sept family of the Síl Muiredaig, and who were transplanted from their original homeland of Tír Briúin na Sinna

8

(McCotter, 2014, pp. 208, 210) into their new territory in Tír Maine, certainly seems to indicate deliberate attempts by the Ua Conchobar to curb the power of the O’Kelly, and simultaneously gain control over Tír Maine in the process.

The Acallam na Senórach contains another very interesting description of St. Patrick’s actions at Lough Croan, preceding the occurrence of the aforementioned miracle:

From there they went…to the Speckled Land, now called the Land of Maine, and to Loch Linngáeth, now called Lough Croan. The King of Connaught, Muiredach Mór ‘the Great’, son of Finnachta, was waiting there for Patrick. A tent was set up for Patrick and his clerics, and the nobles of the province of Connaught approached and did homage to Patrick, putting their heads on his lap.

Patrick then left the tent and sat on an earthen mound…

‘Well, good Caílte,’ said Patrick, ‘whose grave lies in the hill on which we are standing?’ ‘It is the grave of a warrior of the Fíana of Ireland who died here,’ said Caílte, ‘Airnélach, who was the son of Admallán, King of Leinster…This green-clad mound was raised over him, and the stone against which, holy Patrick, you now rest your back was placed above his head.’ Caílte then recited the following poem:

‘The grave under Lodán’s stone, the grave of a man of proud birth, small grave of a glorious soldier, Airnélach, son of Admallán.’ (Dooley & Roe, 2008, pp. 33-35)

The Lía Lodain, as it is described in Irish, tops a burial mound in the narrative, and while caution is required in suggesting a correlation with this literary account and the physical landscape, one speculative location could be offered for these monuments. It is possible that the standing stone mentioned previously, now replaced at the top of ‘Fair Hill’ in the townland of Gortaphuill by a wayside cross, may have been the location being described in this sequence of the Accallam. Further weight can be added to the claim when we consider that the adjacent townland to the south carries the very persuasive name of Tullyneeny.

The described episode refers to Patrick receiving homage from the nobles of Connacht in his nearby tent, an action not dissimilar to elements that are known to have occurred at communal assemblies and inauguration ceremonies recorded from high and late medieval Ireland. With ‘Fair Hill’ overlooking Lough Croan, located 2 km to the north, and with this low hill (73m OD) commanding views over the now turlough, the precinct around the townlands of Gortaphuill and Tullyneeny could, in fact, be held up as a potential undocumented assembly and inauguration venue for the O’Kelly lords of Uí Maine.

Royal lands are normally the preferred setting for annual óenach assemblies (O'Sullivan, 2004, p. 80), and we know from early medieval sources that it was the king’s duty to hold the regular assembly on ‘king’s land’ (Kelly, 2016, p. 403), which would add weight to Lough Croan and its environs being a cenn áit of the Uí Maine and O’Kelly through time.

While it is often assumed that óenach assemblies routinely took place at locations with height viewpoints, there are a number of instances of such assemblies being set at lakes or rivers (McNeill, 2008, p. 67). Moreover, in a number of instances, the location of an óenach is complemented by its proximity to accessible shallow bodies of water, or in the sward vacated by the water in seasonal lakes. There is a strong connection between the symbolism inherent in the fertile summer grasses found in the turlough beds, and the bounty and fecundity of the land that is celebrated at the harvest festival of Lughnasa. This motif could be present at Lough Croan, and the perennial

9 retention of water at Lough Croan in the summer may have been utilised for the swimming of animals, a common purification tradition at these sites (O'Flaherty, 2014, p. 13).

Finally, and returning briefly to the account of Patrick’s visit to Lough Croan in the Acallam na Senórach, the author of the Acallam is careful to describe the events of Patrick’s time there, from the tribute provided to Patrick at this place, through to the significance of including acknowledgement of an antique funerary landmark in the region. There was also, however, careful description of the events of a game of hurling. All three aspects of this account preserve evidence of medieval Irish assembly and possibly inauguration practices, with the performance of games and horse-racing another key element of assembly practices (Gleeson, Kingdoms, Communities, and Óenaig: Assembly Practices in their Northwest European Context, 2015, p. 35). Combined, they may be pointing towards a key location within the precincts of Lough Croan, which was set aside for one of the key duties of a medieval Gaelic king.

The next possible Lough Croan reference occurs in Leabhar Ua Maine for the year 1268. The place of death of Maine son of Conchubhar O’Kelly, king of Uí Maine, is recorded at Gráinseach Chairn Bhuaileadh (Ó Muraíle, 2010, p. 64). As mentioned earlier, the location of Gráinseach Chairn Bhuaileadh can only be speculated at, but there is a possibility that it refers to the townland of Grange (An Gráinseach), located directly to the northeast of the turlough. There are seven ringforts in the townland, one of which may have served as residence and place of death for Maine O’Kelly.

We must wait until the beginning of the fifteenth century for the next historical reference to Lough Croan, and this corresponds with the record of death of Conchubar son of Mael O’Kelly, king of Uí Maine. Conchubar is described explicitly in Leabhar Ua Maine as dying at Loch Cróine (Ó Muraíle, 2010, p. 67). This death record likely refers to either a residence on the lake, or by the lakeshore of Lough Croan, and with the abundance of medieval settlement archaeology in the vicinity, a number of candidates are possible.

The final references to O’Kelly settlement at Lough Croan occurs in 1536 (Freeman (ed.), 1944; 1977, pp. 697-699; Hennessy (ed. & trans.), 1939, pp. 301-303) and 1545 (Hennessy (ed. & trans.), 1939, pp. 347-349), and concern Turrock castle. Located in the townland of the same name, Turrock castle was the possession of Donnchad son of Emann O’Kelly, king of Uí Maine until c.1557. As stated previously, there is likely to be a deliberate link between the castle at Turrock, and the crannóg site of Edward’s Island. Artefactual assemblages from Lough Croan There is a significant artefactual assemblage in the care of the National Museum of Ireland (NMI) for Lough Croan. A total of twenty four items form the collection on record, and twenty two of these items are concentrated on two locations. The majority of these items were found through illegal metal-detecting exploration, and they are recorded as being discovered at the crannóg named Edward’s Island (RO044-107----), and at the redundant record at Illaunamona (RO044-091----). The collection at Edward’s Island comprises ten items: one Jews harp (NMI 1991:23), one object – possibly a tool fragment (NMI 1991:22), one knife with a curved blade (1991:21) [all iron], one decorative flat round button (NMI 1991:20), one object – possibly a gun fragment (NMI 1991:19), one ingot (NMI 1991:18), two decorative mounts (NMI 1991:17 & 1991:14), two trapezoidal buckles (NMI 1991:16 & 1991:15), and one decorative annular brooch-pin (NMI E 499:40 – Not a part of the collection donated in 1991) [all copper alloy or bronze].

One of the copper alloy mounts, (NMI 1991:14), is described by Griffin Murray as a twelfth-century ‘clasp’ with red and yellow champlevé enamel work, and he has deduced that it was probably

10 produced in the same Roscommon workshop that produced the Cross of Cong (Murray, 2014, pp. 138-139).

The assemblage from Illaunamona Island comprises eleven items: one nail (NMI 1991:33), one tool – possible awl fragment (NMI 1991:32), and one hook (NMI 1991: 31) [all iron], one small lead vessel with looped handles (NMI 1991:30), one decorated vessel rim fragment (NMI 1991:26), one decorated disc (NMI 1991:27), one decorated square button (NMI 1991:25), one decorated mount with floral design (NMI 1991:24), and two unidentified objects (NMI 1991:29 & 1991:28) [all copper alloy or bronze].

Aside from these artefactual assemblages, which were all, bar one, donated in 1991, there are two other items which complete the extant collection on Lough Croan. One is a bone pin or needle (NMI 1977:2350), which was found in Garrynphort townland, in a location described as ‘A Crannóg on Lough Croan’. The second is a carved stone head (NMI 1971:952), which was found in Coolnageer townland. Etienne Rynne published the carved stone head, identified it as a king, and surmised it to be part of a king-bishop-queen unit, usually found in late medieval Gothic churches. This carved head seems have served as the right jamb of a window within a church. Rynne also suggested its place of origin may have been a burial ground and ecclesiastical enclosure (RO045-167001- & RO045-167002-) known locally as Caltragh (Rynne, 1966-1971, pp. 92-93), which is located c. 2.2 km east of the eastern shoreline of Lough Croan turlough. There are no church remains extant at Caltragh, but the available evidence on site would suggest that this was once the case.

While this assemblage is quite extensive for an unexcavated archaeological landscape in Co. Roscommon, it is difficult to interpret. The discoveries are, in the main, the result of metal detecting, and does not provide a representative sample selection of the material culture surviving in these locations. Another issue with this collection is the lack of a reliable find circumstance for any of these artefacts, despite the information on file for the collection. Taking into account these limitations, there are still conclusions to be drawn from the available evidence.

If we accept that Edward’s Island and Illaunamona Island are the genuine find locations for these items, then it confirms that these two locations were inhabited in past times. The items seem to indicate that the crannóga were inhabited in the medieval period. Combining the assemblage, and evaluating the material composition of the largest group (Copper-alloy/bronze, 14/24 – 58%), leads us to argue that the character of these items portrays elite habitation, with the presence of a number of decorated copper alloy and bronze artefacts mixed in with the everyday tools and implements. Comber highlights the significance of bronze as a high-status metal in a medieval context (Comber, 2018, p. 100), and such a conclusion can also be advocated for in respect of Lough Croan. The discovery of the possible gun fragment (NMI 1991:19) also points towards later medieval occupation or use of the crannóg of Edward’s Island, considering as the introduction of firearms to Ireland seems to occur in the late fifteenth century (Hayes-McCoy, 1938, p. 47). The discovery of the enamelled clasp (NMI 1991:14) can date the crannóg’s use to at least as early as the twelfth century.

The site of Illaunamona Island is more difficult to interpret from the artefactual assemblage, due to an uncertainty over its location. This author has encountered an inconsistency locally over the location of Illaunamona Island, with local information advocating for a different location to that of the island labelled as such in the First Edition 6-inch Ordnance Survey map for the area. Irrespective of this, the items uncovered suggest settlement activity, and the working of material for making clothes, as well as the personal ornamentation provided by the mounts and decorated disc.

11

Focussed archaeological research at Lough Croan Through the cataloguing of this extensive body of diverse evidence, this author came to the conclusion that the Lough Croan landscape was worthy of further archaeological investigation, and this was initially undertaken via the acquisition of 18 km² of processed Light Detection and Ranging (LiDAR) data, at a posting of one reading per 0.5m, from the Ordnance Survey Ireland (OSi), thanks to a Royal Irish Academy Archaeology Research Grants, funded in 2018. This data was used to create a Digital Terrain Model (DTM) of Lough Croan turlough and its immediate environs.

The acquisition of this topographical data has enabled this author to garner an extremely valuable insight into the Lough Croan landscape. It has provided highly accurate dimensions of the turlough, as well as uncovering links between monuments that an aerial photograph or cartographic source is unable to provide. One of the key questions which this topographical data has been able to address involves the digital reflooding of the turlough of Lough Croan, in order to get an idea of the precincts of the former lake shore, as well as ascertain the true extents of the collection of crannóga that survive on the turlough, in terms of size, shape above the water level, and intervisibility.

The software utilised in digitally modelling water level to 69m OD revealed during its processing a natural peninsula, promontory or natural island, jutting out into the then lake (X: 587885; Y: 749494 ITM). This natural promontory is also visible on the Cambridge University Collection of Aerial Photography (CUCAP) photograph [APH034], along with other features, some of which can no longer be identified on the ground through fieldwalking. This promontory has been described locally as Illaunamona, which would be contrary to the cartographic information available for the area. This also throws some doubt on the find spot for the assemblage of artefacts donated in 1991 for the named location.

Figure 2 - Digital modelling of the water level of Lough Croan turlough to 69m OD. Note the range of 'islands' located on the lake, particularly the collection of landmarks at the centre of the lake. The red arrow indicates the location targeted for remote sensing investigation. (Data courtesy of OSi, image generated by LGS)

This promontory was deemed worthy of reconnaissance, and this author noted that along with the elevation of the promontory above the turlough bed, 2.5-3.5m of a height difference, the perimeter of the summit of the promontory was marked by a circle of boulders, c. 25m in diameter, with further, higher concentrations of stones in adjacent areas also. While there is no RMP ascribed to

12 this feature, the nearest RMP is located 107m to the west of a monument classified as a crannóg in the RMP as RO044-089----. Its description is as follows:

‘Marked as a small island on the 1915 ed. of the OS 6-inch map, and situated on a rise at the SW shore of the turlough Lough Croan (dims c. 3 km NW-SE; c. 200-300m NE-SW). This is a circular spread of boulders (diam. 25m N-S; 25m E-W) on a grass-covered rise (H 0.45m) with some distinctive boulders on the perimeter, but there is no clear evidence of artificial construction. (RO044-088----) is c. 100m to the N.’

Despite this description however, there is no evidence of the aforementioned monument in the area that are described in the co-ordinates.

Figure 3 - Location of area chosen for investigation, outlined in red on a six-inch OSi map with plotted location of RMP RO044-089--- to the west. (Map OSi/ASI)

The plotted location of the site lies some 107m west from the centre of the mound which is more evident in the LiDAR image given in Fig 4 and aerial photography in Fig 5.

13

Figure 4 - LiDAR image of the survey area, marked by a rectangle (LiDAR data courtesy of OSi, acquired courtesy of RIA); Figure 5 - Location of area considered for investigation, with mound outlined in red on an aerial photograph, with plotted location of RMP RO044-089--- to the west. (Photo: Digital Globe)

Figure 6 - Contoured topographical plan of the target location, derived from LiDAR data (Data courtesy of OSi, image generated by LGS)

The geophysical survey sought to identify any structural and superficial components on the mound which may confirm it is a crannóg.

14

1) Survey Control

10m x 10m survey grids were set out over the survey area (See Fig 8) using a combination of a Trimble Ag132 Sub-Metre differential GPS (dGPS), a Sokkia Set 500 total station and measuring tapes.

2) Reconnaissance Survey of Groundfast Stones

This was carried out using a Sokkia Set 500 total station to map the location of visible stones.

3) Magnetic Susceptibility Survey

This reconnaissance survey was carried out on a 2m x 2m grid using a Bartington MS2 Susceptibility Meter with MS2D fieldloop.

4) Earth Resistance Survey

The survey was carried out on a 0.5m x 0.5m grid using a TR/CIA Mk 2 resistance meter and 0.5m twin-probe array.

5) Electrical Resistivity Tomography (ERT)

The survey was carried out on a transect across the mound using a Campus Instruments Geopulse resistance meter controlled by a Toughbook CF-27 laptop computer. A 2m electrode spacing expanded to 6 levels allowed for a modelled pseudo-section to a 6m depth to be created.

6) Data Processing, Visualization, Interpretation and Presentation

Field data was processed using the equipment manufactures’ propriety software. The processed data was visualized using appropriate software packages such as Global Mapper and Surfer. Anomalies due to possible archaeological and non-archaeological sources have been identified, interpreted and described in maps in this report.

Figure 7 - Earth Resistance Survey in progress at Lough Croan, with Total Station in foreground. (Image courtesy of Kevin Barton)

The decision was made to implement a phased investigation of the targeted area at Lough Croan. The rationale for this approach was due to the absence of a recorded monument for the location in

15 question, despite its morphology, which indicates the area is archaeologically prospective. Field visits to the site provided a better understanding of the conditions. The promontory was covered in thick grassy hummocks, with much of this grass growing over substantial quantities of scattered stone. The site was overgrown in places, with vegetation taking hold in some areas of the site. Due to the quantities of scattered stone evident on the promontory, the first survey conducted was a reconnaissance survey of the groundfast stones extant on the summit of the promontory, in order to ascertain if there was any pattern to this arrangement of stones.

Figure 8 - Reconnaissance survey geo-referencing the groundfast stones on the summit of the promontory at Lough Croan, set within the survey grids. (LiDAR data courtesy of OSi, image generated by LGS)

This reconnaissance survey confirmed the existence of a sub-circular arrangement of groundfast stones on the summit of the promontory, with a diameter (c. 25m) similar to that described for the NMS record for RO044-089----.

This survey was followed up by a 100m long Electrical Resistivity Tomography survey, located on a SSW to NNE axis, in order to bisect the summit of the promontory, the location of which can be seen

16 in Fig 8. This technique acquires resistivity data that is modelled to produce a pseudo-section of the relative resistivity variation beneath the ground surface to a depth of c. 6m.

Figure 9 - Electrical Resistivity Tomography (ERT) survey conducted over the summit of the promontory at Lough Croan. (Image courtesy of LGS)

This survey recorded a relatively dense, compacted core of higher resistivity material over a large section of the promontory, beginning at c.32m and continuing until c.88m along the survey line (Fig 9). The depth of investigation (c. 6m) highlights that this core of dense material persists with depth throughout the survey section. This may indicate that this promontory is formed from an outcrop of limestone bedrock or a stony moraine located in the linear depression that forms the lake.

The combination of the Total Station and the ERT surveys informed the application of further techniques to the promontory. The project team decided to follow the initial surveys with an earth resistance and magnetic susceptibility survey over 190m² of the promontory summit.

The earth resistance survey uncovered a number of anomalies within the survey area (Fig 10). The technique has a depth of investigation of c. 0.5 – 0.75m, which can reveal superficial high and low resistance features just beneath the surface. The most prominent of the uncovered anomalies is a distinctive complex circular feature with a diameter of c. 25m, in the western half of the survey area. This is accompanied to the east of the circular feature by an alternate banding of high and then low resistance values running in a north-south direction towards the eastern end of the survey area. The low resistance band to the northeast of the survey area may be evidence of a surrounding the site.

17

Figure 10 - Earth Resistance Survey conducted in 10m x 10m grids. (Image courtesy of LGS)

A magnetic susceptibility survey can reveal areas of activities such as hearths or industrial processes, due to the enhanced susceptibility values that are consistent with repeated burning. The magnetic susceptibility survey at Lough Croan was conducted over the same footprint as the earth resistance survey grids. This investigation revealed a pair of higher susceptibility zones within the 25m x 25m circular anomaly of the enclosure revealed by the earth resistance survey. Should these anomalies be the result of artificial processes, i.e. a hearth fire, then this unrecorded feature may in fact be a settlement site. Fig 11 indicates the high susceptibility values within the survey area, concentrated inside the same zone as the circular feature revealed in the earth resistance survey.

18

Figure 11 - Magnetic Susceptibility Survey conducted in 10m x 10m grids. (Image courtesy of LGS)

The multi-method remote sensing investigations conducted on this location has revealed a variety of evidence from different sources, which may prove to be consistent with settlement and/or industrial activity in this part of the Lough Croan landscape.

Discussion of the Preliminary Results of the Remote Sensing Investigations The phased investigation of this promontory on the southern shore of the turlough at Lough Croan was informed by a number of independent sources of information. Historical sources provided a general understanding that the Lough Croan area was of particular importance to early medieval Uí Maine, an importance that remained for the high and late medieval O’Kelly dynasts of Uí Maine, up until the sixteenth century and beyond. Combining this with the placename evidence surviving for the area has served to provide some insight into the societies who settled and utilised this landscape in the past. The artefactual assemblage extant for the Lough Croan area has been revealing in that it

19 has given some idea as to the strata of medieval society which may have inhabited the crannóga on the lake in particular.

The decision to choose the site targeted for investigation was carefully researched. Firstly, the physical appearance of the promontory, indicated from field visits, aerial photography, and the processing of LiDAR data for the area, provided the author with an indication that this location may have been chosen for lakeland settlement at some point in the past. The presence of a NMS record of strikingly similar description, located 107m to the west of the target site, and with the absence of the described features at the location cited, added further weight to a hypothesis that there may have been a mistake in the geo-referencing of this monument.

Secondly, modern cartographical sources have recorded a series of island names, many of which are still evident on the former lake, which provide further weight to the argument that this landscape was used for lakeland settlement. Local information, and the collection of artefacts reputedly discovered at Illaunamona, which the target site has been called by locals, leads this author to believe that this may, in fact, have been the location targeted by metal-detectorists prior to 1991, an endeavour which yielded eleven items.

The first survey geo-referenced a large collection of groundfast stones on the summit of the promontory (Fig 8), which produced a sub-circular shape, suggesting that the possible foundations of a substantial stone structure, c. 25m in diameter, crowning this promontory. The Electrical Resistivity Tomography survey (Fig 9) strongly suggests that this promontory is natural in composition, and developed the argument that if this location was used for settlement activity, the site was very deliberately chosen for the purpose.

Armed with this information, the earth resistance survey then revealed a substantial complex circular feature in the western half of the survey area (Fig 10), again c. 25m in diameter, and with areas of alternating high and low resistance, which may be evidence of the construction of a ditch, dug in order to provide greater defence to the natural promontory, and the structure within.

The magnetic susceptibility survey (Fig 11) then produced two zones of high susceptibility values which may be consistent with repeated burning activities. These signatures are often the result of activity such a settlement fires or industrial activities such as metalworking, once more implying that this promontory was in use at some point in the past.

Gathering this information together, and noting that what is being interpreted is only preliminary results, the picture that begins to emerge for this promontory is that it may have served as a defended dryland residence or crannóg site, which utilised the naturally formed high ground as a foundation upon which to construct a site. Anomalies which may be consistent with settlement fires have been uncovered from the interior of a complex circular feature, while the area surrounding the base of the promontory to the east may have been modified to produce a ditch, in the areas closest to the shoreline proper to the south and east of the promontory. The evidence thus far, not least its dimensions, seems to indicate that the stone scatter on the summit of the promontory may indicate the former presence of a stone fort, or cashel, and the absence of substantial stone remains today may be explained by the abundance of stone in the surrounding field walls in the area. While cashels are routinely described as being a monument type used in the early medieval period, there is growing evidence to support the continued use of cashels, including cashels on crannóga, in parts of into the later medieval period (Comber, 2018; O'Conor, Brady, Connon, & Fidalgo- Romo, 2010). Indeed there is a strong argument to be made for the elite of later medieval Gaelic Ireland actively seeking to portray a style of life and living that is rooted in an antique past, as a

20 means of cementing a legitimacy and power over their territory in the present, particularly in the face of political upheaval (O'Conor K. , 2018, p. 140).

It has been demonstrated that this area was of particular significance to the Uí Maine, and their dominant sept of the O’Kellys, throughout the medieval period, and while is impossible to ascribe particular historical references to a site such as this, there are a number of records within the historical sources which may just refer a high-status residence on the shores of Lough Croan.

The earliest reference, found in the Annála Connacht for the year 1260, describes how a party of Aed O’Conor's followers burned the longport of Conchobar O’Kelly (Freeman (ed.), 1944; 1977, p. 135). The Annals of the Four Masters for the same year reports that a garrison of Conor O'Kelly was burned by the people of Hugh O'Conor (Hennessy (ed. & trans.), 1939, p. 379). Kieran O’Conor referred to this incident in relation to the use of the term longport in the annals, which may be used to describe a stronghold of cashel, ringfort or moated site morphology (O'Conor K. D., 1998, p. 85). It is possible that the longport in question may have been our promontory.

There is an interesting reference in a recently translated praise poem for Uilliam Buide O’Kelly, king of Uí Maine 1349-1381, which begins Tath aenfir ar iath Maineach. The poem, composed c.1345, is used to laud the kingship of Uilliam Buide, and while it does not possess an abundance of locational information relating to Lough Croan, it does include a detailed description (quatrains 14-22) of Uilliam Buide’s residence or court, constructed of wattle, daub and timber. The feast is described as such:

Q19 ‘Many are the players of brandubh playing around its doorposts and its blackbird-thronged wall of shields, the hazel-white hostel whose house-post is not crooked.’

(Hoyne (forthcoming), 2018, p. 40)

The ‘blackbird-thronged wall of shields’ immediately brings to mind the townland name directly adjacent and to the north of Lough Croan - Lisnagavragh – Lios na gCabhrach (the ringfort of the embossed shields). Connecting this description with the placename is purely speculative, but interesting nonetheless.

More than this however, the poem describes Uilliam Buide’s dún as follows:

Q21 ‘Around the fort of the descendant of Cairbre which is surrounded by sea-water, there is no shortage of buxom women swimming on their sides, women swimming on their sides on a wave without pause, while [the wave’s] crest is spray-white because the womenfolk.’

(Hoyne (forthcoming), 2018, p. 42)

This quatrain describes the fort being ‘surrounded by sea-water’, and while there may be an element of artistic licence in the use of the term sál – sea-water, due to the O’Kelly lordship being inland in geographical character, this author believes that this quatrain may in fact be describing a crannóg or natural island residence, still in use by Uilliam Buide O’Kelly in the mid-fourteenth century.

The last historical reference is brief but illuminating. The record of death of Conchubar son of Mael Seachlainn O’Kelly, king of Uí Maine, explicitly describes him in Leabhar Ua Maine as dying at Loch Cróine (Ó Muraíle, 2010, p. 67). This death record likely refers to either a residence on the lake, or by the lakeshore of Lough Croan, and this promontory, on the southern shoreline of the lake, may just be the location intended.

21

Future Work The author would like to continue to work in the Lough Croan landscape beyond the limits of this heritage research bursary, as he believes that continued investigation into this area may provide more key information on what seems to be ancestral homeland of the Uí Maine, and a landscape which is returned to by the O’Kellys generation after generation throughout the medieval period. The author would like to complete investigations on the promontory, with additional earth resistance and magnetic susceptibility grids to the east and south of the current survey area, in order to confirm if a ditch was constructed to surround the site. There are also suitable areas for potential magnetic gradiometry survey in the eastern section of the promontory, which may yield further information.

There are also a number of other locations in the area which could be investigated, not least the link between Turrock castle and Edward’s Island, as well as possible research opportunities in the townlands of Tullyneeny, Cuilleenirwan and Gortaphuill, in order to better understand the relationship between the archaeology in these townlands. Separately, this author will also entertain the possibility of obtaining funding to acquire a sample for radiocarbon dating of the dug-out canoe (RO044-124----), in order to see if it fits into the chronology being studied as part of the PhD research.

22

Acknowledgements The author would like to sincerely thank all who have assisted in the completion of this project. I would first like to thank Nollaig Feeney, Heritage Officer, Roscommon County Council, for granting me the Heritage Research Bursary for 2018. I would like to thank Dean and Pauline Fallon, for graciously allowing access to their land, in order to undertake the survey. I would like to thank Noel Hoare and Mike Gacquin, for facilitating contact with local families, for their knowledge, guidance and assistance. I would like to thank my father Danny Curley, for his help in conducting the survey. Finally, I would like to thank Kevin Barton, Landscape & Geophysical Services, for the equipment, experience and expertise to complete the remote sensing investigations at Lough Croan in a timely and accurate manner, as well as his assistance in interpreting the results.

23

Bibliography Anon. (1904). 36th Report of the deputy keeper of public records, Ireland. Dublin.

Collins, J. F. (2016). Chapter One - Geology, Soils and Cattle Production. In M. O'Connell, F. Kelly, & J. H. McAdam (eds.), Cattle in Ancient and Modern Ireland: Farming Practices, Environment and Economy (pp. 1-11). Newcastle: Cambridge Scholars.

Comber, M. (2018). The tale of items lost...: What objects tell us about the life of a medieval family at Caherconnell cashel, Co. Clare. In J. Fenwick (ed.), Lost and Found III: Rediscovering more of Ireland's Past (pp. 95-102). Dublin: Wordwell Books.

Connon, A. (2014). The Roscommon locus of Acallam na senórach and some thoughts as to tempus and persona. In A. Doyle, & K. Murray, In dialogue with the Agallamh: essays in honour of Seán Ó Coileáin (pp. 21-59). Dublin: Four Courts Press.

DAHG. (2013). Site Name: Lough Croan Turlough SAC - Site Code: 000610. Department of Arts, Heritage and the Gaeltacht.

DCHG. (2008-2018). Retrieved from Placenames Database of Ireland - Logainm.ie: https://www.logainm.ie/en/

Dooley, A., & Roe, H. (2008). Tales of the Elders of Ireland: A new translation of the Acallam na Senórach. Oxford: Oxford University Press.

Doran, L. (2004). 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 , 57-80.

Finan, T. (2010). Medieval Lough Cé. Dublin: Four Courts Press.

Freeman (ed.), A. (1944; 1977). Annála Connacht: the Annals of Connacht, AD 1224-1544. Dublin.

Gleeson, P. (2015). Kingdoms, Communities, and Óenaig: Assembly Practices in their Northwest European Context. Journal of the North Atlantic 8(), 33-51.

Gleeson, P. (2018). Finding the royal heartland of Dál Riada: landscape and kingship in Dunseverick, Co. Antrim. In J. Fenwick (ed.), Lost and Found III: Rediscovering more of Ireland's Past (pp. 181-192). Dublin: Wordwell Books.

Goodwillie, R. N. (1992). Turloughs over 10ha - Vegetation survey and evaluation. Unpublished report to NPWS.

Hayes-McCoy, G. A. (1938). The Early History of Guns in Ireland. Journal of the Galway Archaeological and Historical Society, Vol. 18 No. 1/2, 43-65.

Hennessy (ed. & trans.), W. (1939). The Annals of Loch Cé, RS 54. Dublin.

Hoyne (forthcoming), M. (2018). Uilliam Ó Ceallaigh, Gairm na Nollag 1351 agus stair eacnamaíoch fhilíocht na scol.

Kelleher, J. V. (1971). Uí Maine in the annals and genealogies to 1225. Celtica 9, 61-112.

Kelly, F. (2016). Early Irish Farming: a study based mainly on the law-texts of the 7th and 8th centuries AD. Dundalk: Dublin Institute for Advanced Studies.

24

Logue, P. (2018). All things to all men: Aodh Ó Néill and the construction of identity. In E. Campbell (eds.), E. FitzPatrick, & A. Horning, Becoming and Belonging in Ireland, AD c.1200-1600 - Essays in Identity and Cultural Practice (pp. 269-292). Cork: Cork University Press.

McCotter, P. (2014). Medieval Ireland: Territorial, Political and Economic Divisions. Dublin: Four Courts Press.

McNeill, M. (2008). The Festival of Lughnasa: A study of the survival of the Celtic Festival of the beginning of harvest. Dundalk: University College Dublin.

Murray, G. (2014). The Cross of Cong: A Masterpiece in Medieval Irish Art. Spain: Irish Academic Press.

Ó Lochlainn, C. (1940). Roadways in ancient Ireland. In J. Ryan (ed.), Essays and studies presented to professor Eoin MacNeill on the occasion of his seventieth birthday, May 15th 1938 (pp. 465– 474). Dublin: Three Candles.

Ó Muraíle, N. (2010). The Ó Ceallaigh Rulers of Uí Mhaine - A Genealogical Fragment, c. 1400, Part II. The Journal of the Galway Archaeological and Historical Society 62, 51-77.

Ó Riain, P. (1997). When and why Cothraige was first equated with Patricius? Zeitschrift fur Celtische Philologie 49-50, 698-711.

O'Conor, K. (2018). Crannóga in later medieval Ireland: Continuity and change. In E. Campbell, E. FitzPatrick, & A. Horning (eds.), Becoming and Belonging in Ireland, AD c. 1200-1600 (pp. 148-166). Cork: Cork University Press.

O'Conor, K. D. (1998). The archaeology of medieval rural settlement in Ireland. Dublin: Royal Irish Academy.

O'Conor, K., Brady, N., Connon, A., & Fidalgo-Romo, C. (2010). The Rock of Lough Cé, Co. Roscommon. In T. Finan (ed.), Medieval Lough Cé (pp. 15-40). Dublin: Four Courts Press.

O'Donovan (ed. trans.), J. (1851-1856). Annals of the kingdom of Ireland by the Four Masters. Dublin.

O'Flaherty, E. (2014). Archaeological Watermarks: Turlough Floodplains as Communal Spaces and Places of Assembly. Online Document.

O'Sullivan, C. M. (2004). Hospitality in Medieval Ireland, 900-1500. Dublin: Four Courts Press.

Parkes, M., Meehan, R., & Préteseille, S. (October 2012). The Geological Heritage of Roscommon - An audit of County Geological Sites in Roscommon. Dublin: Irish Geological Heritage Programme, Geological Survey of Ireland.

Rynne, E. (1966-1971). Stone Head from Curraghboy, Co. Roscommon. Journal of the Galway Archaeological and Historical Society, Vol. 32, 92-93.

Stokes (ed. and trans.), W. (1887). The Tripartite life of Patrick : with other documents relating to that saint. London.

25