MEDIEVAL COMMUNICATION ROUTES THROUGH LONGFORD

AND ROSCOMMON AND THEIR ASSOCIATED SETTLEMENTS

By LINDA DORAN*

Received 10 September 2003. Read 16 March 2004. Published 30 September 2004.x

ABSTRACT

This paper explores the direction and context of medieval communication channels

in the territory covered by the modern counties of Longford and Roscommon.

The network consisted of roadways—both local and interregional—and

water-based arteries. The landscape of the area dictated how people moved across

the terrain. Large tracts of bog generated a need for trackways to provide access

to good land trapped in the peat. The extensive water system centred on the

Shannon facilitated travel to otherwise isolated places. The numerous islands contain

the remains of secular and religious settlements. The roads identified as

belonging to the regional network are plotted on a map of the area. This mapping

shows that the main factors shaping the road network were the location of the

ritual centre at Cruachain and the siting of religious establishments. Two major

roads—the Slighe Assail and the Slighe Mho´r—linked an otherwise isolated area to

the east-coast ports and to English and continental markets. The relationship of

the roads in particular to medieval settlement patterns is examined.

Introduction

Roadways are an ever-present but largely invisible feature of the history of medieval

Ireland. Then as now, an adequate communications network was a key ingredient

in social progress, yet we encounter roads in history, as we do today, mainly

through the events known to have taken place on or near them. Roads are mentioned

in different contexts in the medieval records—as the routes of war parties

and cattle raids; in land grants and boundary descriptions; and, on one occasion,

in a dispute sparked when a landowner attempted to incorporate part of the highway

into his own land.1 They are also depicted as sources of danger. In the official

records of the Anglo-Norman administration, control of roadways is vital to the

safety of local communities and the conduct of the king’s business. Official sources

discuss in detail the need to ‘reconnoiter the passes used by the Irish of the

mountains’, and the men of Arst in Co. Kildare were ordered to take a ‘spade

and barrow for two days’ and to cut the roads and entries to the settlement in

order to prevent attacks.2 The exchequer compensated people not only for protecting

the roads and passes, but also for attending to the king’s business in dif-

1ALC, vol. 1, 87y9; Ann. Conn., 33y5; Cal. justic. rolls Ire., 100, 172y3; Civil Survey, vol. 10,

45y9.

2Irish exchequer payments, 302y3; Stat. Ire., 12y22 Edw. IV, 765y7.

58 Proceedings of the Royal Irish Academy

ficult conditions caused by the poor state of the roads.3 The condition of the

highway was also an acceptable excuse for the failure to obey a writ.4 On the other

side of the frontier, the Gaelic Irish annals viewed roadways more neutrally. Death

notices cite the building of toghers and causeways as praiseworthy acts of the

deceased.5 Knowledge of the routeways and passes was seen as an important survival

and strategic skill.6 Furthermore, in a terrain such as this, where good agricultural

land was at a premium and where large tracts of water and bog dominated

the landscape, usable access routes were a fundamental necessity.

Very little evidence remains regarding the extent, condition or location of

roads in medieval Ireland. The annals record that ToirrdelbachO´ Conchobair was

in the field in Ormond for six months between August 1126 and February 1127

but make no reference to the roads travelled.7 Almost no academic research has

been carried out on the subject. The article published by O´ Lochlainn in 1940

on roadways still stands alone as an overview of the island as a whole.8 I have used

the routes isolated by him as the basis for the map that accompanies this text (Fig.

1). It is particularly unfortunate that communication routes are not always listed

in archaeological inventories: they are essentially instruments of settlement and

should be viewed as an element of the settlement pattern.9 Few excavations have

been directed specifically towards the consideration of communication lines. An

obvious and exceptional example is the examination of toghers carried out by the

Irish Archaeological Wetland Unit.10 Since routeways were linked to human habitations,

most have come to light as a result of the archaeological investigation of

an adjacent site, most commonly a rescue excavation or an excavation preceding

development.11

Longford–Roscommon is perhaps a more appropriate geographical area in

which to study medieval communication routes than appears to be the case at first

glance. Although it is remote from the coastal ports and major population centres,

it is crossed by two major routes; it also has a substantial network of minor roads

and toghers. In addition, the region is served by an extensive waterway system

based on a number of navigable rivers, most notably the Shannon, along with its

tributaries and lakes (Fig. 1).

3Irish exchequer payments, 474, 302y3; James Graves (ed.), A roll of the proceedings of the king’s

council in Ireland, for a portion of the sixteenth year of the reign of Richard the Second, A.D. 1392y93

(London, 1877), 187.

4Graves, Roll of the proceedings of the king’s council in Ireland, 189y90.

5Ann. Conn., 493.

6Ann. Conn., 35.

7F.J. Byrne, ‘The trembling sod: Ireland in 1169’, in T.W. Moody et al. (eds), A new histor y of

Ireland. Vol. 2: Medieval Ireland, 1169–1534 (Oxford, 1987), 1–42: 11.

8Colm O´ Lochlainn, ‘Roadways in ancient Ireland’, in John Ryan (ed.), Fe´il-sgrı´bhinn Eo´in

Mhic Ne´ill: essays and studies presented to Professor Eo´in Mhic Ne´ill on the occasion of his seventieth birthday

(Dublin, 1940), 465–74.

9There is, however, a growing tendency to include communication routes in such surveys:

see C.R. Wickham-Jones, The landscape of Scotland (Stroud, 2001); Kieran Jordan and Tom

O’Connor, ‘Archaeological sites of interest surrounding the Turoe Stone’, Journal of the Galway

Archaeological and Historical Society 60 (2003), 110–16: 112–14.

10Transactions of the Irish Archaeological Wetland Unit (Dublin, 1993–).

11For example, a portion of the ‘Red Earl’s Road’ was uncovered as a result of the need to

excavate close to Ballinafad Castle, Co. Sligo, before the construction of a field study centre:

Martin Timoney, ‘Gortalough, Ballinafad’, Excavations 1997 (1998), 155; Excavations 1999 (2000),

272.

DORAN—Medieval Communication Routes 59

FIG. 1—Longford and Roscommon: landscape and communications.

Landscape

The modern counties of Longford and Roscommon are part of the central lowlands

of Ireland. The main rock—carboniferous limestone—is almost entirely concealed

by a deep covering of recent glacial deposits and peat bogs. This glacial

drift is deposited as long gravel ridges, or eskers, which provide routeways through

60 Proceedings of the Royal Irish Academy

the lowland bogs and Shannon callows. The Slighe Mho´r, also known as the Eiscir

Riada, makes use of these gravel ridges (see Fig. 1). In the north and east of the

region, the drift is moulded into steep-sided hillocks known as drumlins; lakes and

bog patches tend to form in the hollows between them, particularly in north

Roscommon. In Longford these drumlins lie along a north-west to south-east axis

in the western portion of the county, appearing as islands of dry land in the areas

of bog. They are important in the settlement of the region, providing patches of

relatively good land throughout the bog. The large number of toghers that run

across these peatlands are concentrated in the drumlins.

Longford and Roscommon recall in their townland names and other place

names the existence of wooded areas no longer extant.12 Roscommon had 6%

woodland cover at the time of the Civil Survey of 1654–6, second only to that

recorded for Clare at 7%. It would appear that the area around Lough Ree was

historically a good locus for semi-natural native woodland. Today, however, only a

few isolated remnants remain, most notably St John’s Wood, Rindown Wood (both

on the Rindown peninsula) and Hare Island (in the southern part of Lough

Ree).13 The inclusion of these woods in the Civil Survey suggests a continuity of

woodland cover since that time. The extent of woodland in County Longford does

not seem to have been as great as that in County Roscommon. A survey of 1618

gives the total acreage of profitable timber for the county as 8,400 acres.14 Wooded

areas were, of course, barriers to travel; in fact the common prefix to road names

was slighe, meaning a highway cut or cleared through woodland.

Both counties are covered by extensive areas of raised, or midland, bog, a

feature that largely determined how the landscape was traversed. The largest continuous

expanse is in Longford, where approximately 20% of the county is covered

by raised and fen bogland. The bog landscape of today differs considerably from

that which existed even a century ago. Although the large-scale drainage and

development schemes that began at the end of the nineteenth century had a

significant effect on the landscape, it was not until the establishment of Bord na

Mo´na in the 1950s that a widespread transformation occurred. Indeed, the extent

of peat cover is not likely to have changed greatly between the twelfth and the

eighteenth centuries. Bord na Mo´na has harvested the bog on an enormous scale:

in County Longford, for example, the total area of operation is 12,000 hectares,

with a similar area in County Roscommon in development. This has had a devastating

effect on the archaeological remains preserved by the bog. The rate of

destruction for sites recorded by the Irish Archaeological Wetland Unit in County

Longford runs at 70%, indicating the fragile nature of this evidence and the

importance of survey.15 Harvesting of the bog had a negative effect on the sites

considered for this study because most medieval routeways would have been closer

to the surface. As well as an important resource, the bog was a barrier, particularly

12In County Roscommon, for example, ‘derry’, meaning an oak wood, is recalled in

Derrycashel and Derrygirraun. Farnbeg indicates alders, whereas Funshinagh denotes ash. In

County Longford tree species are recalled in the following place-name elements: Farnagh (alder),

Cullentragh (holly), Daroge (oak) and Derryaroge (oak grove). The place name Derryaroge

probably recalls the wood from which an adjacent togher was made.

13The Heritage Council, Rindown, Co. Roscommon: a management plan (Kilkenny, 1998), 44.

14J.H. Andrews, Plantation acres (Omagh, 1985), 76 n. 12.

15The Irish Archaeological Wetland Unit was established in 1990 by the Office of Public

Works in conjunction with University College, Dublin.

DORAN—Medieval Communication Routes 61

PL. I—Elfleet tower house overlooking Lough Ree.

to communications. The annals refer to the role of bogs in slowing the progress

of escaping raiding parties.16 The existence of bogs embayed in areas of fertile

ground—making passage from one point to another impossible except by long

detours and, in some cases, cutting off the fertile ground completely—led to the

construction of toghers.17

The most distinctive natural feature within the area is the River Shannon (Fig.

1), which was also the most important communication channel in the region during

the medieval period. In an era when water travel was often safer and more

expedient than moving by road, such an artery was a valuable resource. With large

tracts of bog constituting a barrier to passage, settlement was more directly related

to water than was the case in places with easier access by land. A journey by water,

however, was not always easy or safe. Many rivers could be treacherous, particularly

when in flood.18 Water transport would certainly have been the preferred method

for the movement of goods, especially heavy items such as stone, timber or wine.19

With its tributaries, the Shannon is the main drainage channel of the central

lowlands. The area contains numerous lakes, the largest of which is Lough Ree

(Pl. I). A number of lakes are also found on the major tributaries of the Shannon.

The River Boyle, the principal tributary in the upper reaches of the Shannon,

16Ann. Conn., 237.

17Barry Raftery, ‘A wooden trackway of Iron Age date in Ireland’, Antiquity 60 (1986), 50–3.

18An amusing reminder of the perils of a river voyage is found in Theo Dorgan’s foreword

to T.F. O’Sullivan, Goodly Barrow (Dublin, 2001), 5–7.

19Timothy O’Neill, Merchants and mariners (Dublin, 1987), 55–6; Christopher Norton, ‘The

export of decorated floor tiles from Normandy’, in Jenny Stratford (ed.), Medieval art, architecture

and archaeology at Rouen (London, 1993), 81–97: 81.

62 Proceedings of the Royal Irish Academy

PL. II—Carraig Mac Diarmata, Co. Roscommon.

flows out of Lough Gara and through Lough Key, a shallow lake with many substantial

wooded islands. Some of these islands have important archaeological

remains: the monastery of Trinity Island, for example, and the remains of the Mac

Diarmata stronghold at Carraig Mac Diarmata (Pl. II). Cranno´gs are found in the

lakes of both counties under discussion. Evidence, both archaeological and documentary,

for the use of cranno´gs during the later medieval period is plentiful.

Certain sites, such as Ardakillin and Kilglass in County Roscommon, are mentioned

repeatedly in the medieval annals. Ruaidhri O´ Conchobair’s house at his

longphort of Ardakillin Lough, for instance, is cited a number of times by the

annalists, who record a variety of events across a broad chronological span.20 Excavations

at Lough Kinale, Co. Longford, produced not only the fragment of a

decorated book-shrine, but also a distinctive lozenge-shaped stickpin, probably

from the twelfth century. In addition, the palisade gave a radiocarbon date of the

first half of the twelfth century.21

Island monasteries such as Trinity Island are a significant feature of settlement

in these two counties. They would have played an influential role in the development

of a communications network. The interrelation between these religious

houses and the surrounding social and physical landscape is a vital and underresearched

topic.22 The island monasteries of Lough Ree in the south and Lough

20Ann. Conn., 189, 538, 565; ALC, vol. 1, 79; Aidan O’Sullivan, The archaeology of lake settlement

in Ireland (Dublin, 1998), 152–6.

21E.P. Kelly and Nessa O’Connor, ‘Lough Kinale, Tonymore North’, Excavations 1987 (1988),

22.

22For a discussion on the subject, see O’Sullivan, Archaeology of lake settlement, 147–9, 161–3.

DORAN—Medieval Communication Routes 63

Key in the north must have dominated these waterways in medieval times. Many

were important economic and pilgrimage centres; placed on sailing routeways,

they had ready access to the large area around the lakes. Long-distance shipping

moving up and down the country by means of the rivers and lakes possibly used

these monasteries and their related shoreline houses as hospitals, markets and

sources of food. Church settlements at river crossings could also have served as

hostels for pilgrims moving along the waterways. The monastery at Athleague in

County Roscommon, for example, overlooked an important ford on the River

Suck. The market settlement founded in 1231 at Rockingham on the shore of

Lough Key was no doubt also related to water traffic on the lake. The importance

of water transport and travel in this area is illustrated by the arduous efforts made

by the crown to control Lough Ree—exorbitant amounts were spent to ensure

that the royal castle and settlement, sited on the lake at Rindown, could be supplied

and reinforced.23

Broad callows extend on either side of the River Suck in County Roscommon

and on either side of the River Shannon around Lanesborough and south of

Athlone. These are flat fields or marshland in the floodplain of the river. They

are inundated in the winter and slowly dry out as the river subsides in the summer.

Because the callows receive several silt-carrying floods in the year, the soil is particularly

rich and is an important element in the agriculture of the region. Excavations

revealed a small sand-and-gravel road leading from the callows towards the

old burial ground at Clonmacnoise.24 On the whole, the Shannon callows are

relatively narrow, at most points no more than 800m wide; however, those at Cloonburren

are more than 1.6km across. These callows have made a broad river more

difficult to ford or bridge.25

Fording places and bridges

The Sites and Monuments Record for County Roscommon notes 83 fords and 63

stepping stones; for County Longford the same survey records 25 fords. O’Keeffe

and Simington suggest that stepping stones are frequently found at old surviving

fords.26 Joyce observes that in English transliterations of Irish place names, the

most commonly used word elements associated with fords are ‘at’, ‘ah’, ‘ash’ and

‘a’ (all derived from the Irish a´ th).27 Sometimes these are combined with word

elements such as ‘bally’ (baile, ‘township’) or ‘bel’ (be´al, ‘mouth’) to denote the

settlement that grew up around the ford, as happened at Be´al A´ tha Liag (Lanesborough).

These elements can occur as a prefix or as part of the name, as in

Aghabehy, Tullaghmore and Bella in County Roscommon. This form of place

name commonly describes a feature that defines a barony boundary, for example

Belacloghan, Co. Longford.28 Sometimes these names are qualified by an adjective

23Sheelagh Harbison, ‘Rindown castle: a royal fortress in Co. Roscommon’, Journal of the

Galway Archaeological and Historical Society 47 (1995), 138–48: 144.

24H.A. King, ‘Clonmacnoise monastery, new graveyard’, Medieval Archaeology 42 (1998), 169.

25Stephen Heery, The Shannon floodlands: a natural histor y of the Shannon callows (Kinvara,

Galway, 1993), 18.

26Peter O’Keeffe and Tom Simington, Irish stone bridges: histor y and heritage (Dublin, 1991),

25.

27P.W. Joyce, The origins and histor y of Irish names of place (Dublin, 1875), 24.

28Civil Survey, vol. 10, 48.

64 Proceedings of the Royal Irish Academy

PL. III—Pallas Ford, Co. Longford.

giving extra information on the characteristics of the ford or the topography of

the surrounding area. One of the best examples of this is Athleague, Co. Roscommon—

A´ th Liag, meaning ‘the ford of the stones’.29 Some place names recall an

activity that took place at the ford. The small island of Sna´mh Da´ E´an (Swim Two

Birds), close to Cloonburren, probably denotes a place where livestock could be

crossed safely.

In an era of few roads and fewer bridges, fords were vital arteries of communication.

They were an integral part of daily travel across the landscape. They were

essential, for instance, for moving animals across the countryside, and many had

been in use for generations. Control of fording places was necessary to any successful

military campaign, and some, such as Be´al A´ tha Liag and Termonbarry,

became the focus for settlements.

The annals contain a number of references to meetings and battles at fords.30

Many of the latter took place when retreating armies were caught by pursuers—

the Annals of Connacht, for example, record the killing of Tomas O´ Fearghail in

1462 at Pallas Ford, Co. Longford, by a party of Dillons in alliance with another

branch of the O´ Fearghail (Pl. III).31 The ford at Be´al A´ tha Liag, where there was

later a bridge, features in one of the first Anglo-Norman forays into Connacht. In

1177 Miles de Cogan advanced into Roscommon with a regular force of 40 knights,

200 mounted archers and 300 foot soldiers. Establishing a pattern that was to mark

almost all the Anglo-Norman military adventures in Connacht, they were aided by

29Deirdre Flanagan and Laurence Flanagan (eds), Irish place names (Dublin, 1994), 19.

30Ann. Conn., 113, 243–5, 355; ALC, vol. 1, 407.

31Ann. Conn., 511.

DORAN—Medieval Communication Routes 65

PL. IV—Correenbeg Ford, Co. Roscommon.

a disaffected member of the O´ Conchobair. Murchad, the son of Ruaidhri (who

at this time was making a royal circuit of west Connacht), acted as their guide. De

Cogan initially had some success, and he and his force advanced as far as Tuam,

the ecclesiastical capital of Connacht, where they spent three nights. On hearing

of Ruaidhri’s advance they retreated across the Shannon by the ford at A´ th Liag,

but Ruaidhri arrived before they could make good their escape. The Annals of

Tigernach record that the ‘panic at the ford was painful’.32 In 1266 the same annals

record that Mac William Burke, returning from a hosting against Ua Ma´el

Sechlainn, lost many of his men to drowning at the ford near Shannonbridge.33

In 1990 an archaeological search of the river bed at Correenbeg Ford on the

Suck in County Roscommon yielded a rich assemblage of artefacts from the Neolithic

to recent times (Pl. IV). The search also identified an enclosure, a castle

and possibly a cranno´g close to this ford.34 On the Roscommon bank of the river,

an earthwork—the remains of a defensive structure protecting the ford—was

surveyed.

From earliest times the gateway to and from the west lay across the Cloonburren

callows, close to the later motte. Here a pilgrim road runs east–west across

Ireland along eskers via Clonmacnoise, crossing the Shannon at the small island

32Ann. Tig., 299–300.

33Ann. Conn., 147.

34SMR RO056-02701/2; E.P. Kelly, ‘Correen Ford’, Medieval Archaeology 34 (1990), 237.

66 Proceedings of the Royal Irish Academy

of Sna´mh Da´ E´ an.35 In prehistoric times this may have been the only crossing

point between Athlone and Lough Derg, and in the medieval period it was still a

major fording point.36 The Irish Underwater Archaeological Research Team discovered

the remains of a timber bridge to the north (Fig. 1).37 The Annals of

Clonmacnoise mention a bridge at the site in 1158. During the Wetland Unit survey

of the Blackwater area of County Offaly, a substantial gravel road was uncovered

in Coolumber Bog. This road still provides the foundations for a modern road

from Nure, Co. Roscommon, until it enters Cloonburren Bog from the north-west.

It runs directly towards Clonmacnoise on the opposite side of the Shannon. It

may have been an access road to the monastic site; unfortunately, its date and

location to the east of the bog are unknown. Local tradition records a funeral

route that passed through Nure from the early monastic site and St Brigid’s holy

well at Drum to Clonmacnoise.38 FollowingO´ Lochlainn, this road may have linked

up with the Slighe Mho´r to the north of Cloonburren. Investigations of the bridge

by the Irish Underwater Archaeological Research Team suggest an early ninthcentury

date. It is therefore the earliest known bridge in Ireland and also the

largest free-standing wooden structure in early medieval Ireland. It was constructed

of at least fifty oak posts, each up to 15m long, driven to depths of 3–4m into

the soft silts of the river bed. This was a significant achievement and probably

required pile-driving of the posts from rafts into the river bed. Remarkably, it

seems likely that the level of the present river bed is fairly close to that of the

early medieval period. No evidence has been found for repairs to this bridge or

for the construction of a second one at this place. Given that the lifespan of

exposed timbers is generally fifty years, it is likely that the bridge went out of use

some time in the mid-ninth century.39

Although this bridge was short-lived, it is important for a number of reasons.

It is remarkably early—most documentary references to bridges date from the

eleventh century at the earliest. It was a considerable feat of engineering and

would have required the command of a wide range of resources.40 This was a

frontier region for all of the medieval period, and during the late eighth and early

ninth century, political tension was heightened in the midlands. The political

cooperation needed, therefore, to organise this construction project was significant.

The perceived commercial and social rewards must have been considerable.

Clonmacnoise, as well as being a major ecclesiastical centre, was the site of one

of the three chief fairs of Ireland. Fairs seem to have been held in Ireland from

about A.D. 800 onwards.41 The reason for the destruction of the bridge is

35Togher no. RO-CLN 0001 in the Blackwater survey may be the part of the routeway, see

Transactions of the Irish Archaeological Wetland Unit 4 (1995), 130. O´ Lochlainn suggested that the

name may refer to a place where horses or cattle could swim across the river without danger:

Lochlainn, ‘Roadways in ancient Ireland’, 466.

36Ann. Tig., 155.

37Fionnbarr Moore, ‘Ireland’s oldest bridge—at Clonmacnoise’, Archaeology Ireland 10 (1996),

24–7.

38Edward Egan, ‘Ancient funeral route to Clonmacnoise’, Journal of the Roscommon

Archaeological and Historical Society 6 (1996), 70–1.

39Tom Condit and Gabriel Cooney (eds), The Clonmacnoise bridge: an early medieval river crossing

in County Offaly, Heritage Guide (Dublin, 2000), unpaginated w5x.

40I am grateful to Mr Bill Doran for advice on the engineering techniques involved in the

building of this bridge.

41Condit and Cooney, Clonmacnoise bridge, w5x.

DORAN—Medieval Communication Routes 67

unknown; however, it may not have been an engineering failure. Viking activity

on the Shannon may have led to the destruction of the bridge or rendered it a

liability. This location had always been a major crossing point and continued to

be so even after the bridge passed out of use.42

The trackway in Coolumber Bog is directly aligned with the western end of

the bridge and may have been built at the same period or, since it could also have

linked to a ferry, may pre-date the bridge.43 This hints at the existence of a settlement

that was linked to the crossing point at the Cloonburren side of the river.

Recent excavations within the monastic enclosure at Clonmacnoise show that

between the late eighth and the early ninth centuries the monastic settlement was

expanding, with house structures and industrial working areas spreading down

towards the riverbank.44 For a growing population associated with the monastery,

the bridge provided access to farmland across the river. It also allowed people,

livestock and goods to travel to markets and fairs held at Clonmacnoise. From the

east the bridge could have been approached dry-shod, but on the Roscommon

side the traveller faced miles of almost impenetrable raised bog served by toghers.

O’Donovan referred to the Coolumber/Cloonburren road in the Ordnance

Survey letters, mentioning the existence of a cross in the middle of the routeway

in addition to a holy well, which had dried up as a result of the digging of drainage

channels on either side of the road to reduce flooding. A cross base with an

accompanying well is marked on the Sites and Monuments Record for County

Roscommon. Crosses are frequently associated with medieval roadways.45 Tracing

the line of the road from the nearest western dryland to the bank of the river

indicates that over 2.2km of the Coolumber road would have been constructed

over the bog. As with the Bloomhill road excavated in County Offaly, the gravel

and clay used in the Cloonburren road were obtained locally from glacial deposits;

nevertheless, its construction would have required considerable effort and

organisation.46

The underwater surveys and excavation uncovered a range of finds on the

river bed.47 Up to eleven fragments of dug-out wooden boats were discovered. Two

were found lying beside timbers of the bridge, while others were located both

upstream and downstream.48 These boats, originally carved from whole roundwood

trunks with axe and adze, vary in form. Such craft may have been used for

ferrying passengers or transporting goods across the river; some dug-outs showed

evidence of repair. Since they were found close to the bridge timbers, they are

presumably of similar date. During a survey of Lough Gara in the 1950s, seventeen

42I am indebted to Dr Howard Clarke for discussion on the historical and economic context

of the bridge at Clonmacnoise.

43Transactions of the Irish Archaeological Wetland Unit 4 (1995), 5, 130; Heery, The Shannon

floodlands, 18–20.

44H.A. King, ‘Excavations at Clonmacnoise’, Archaeology Ireland 6 (1992), 12–14.

45For discussion, see H.A. King, ‘Late medieval crosses in county Meath, c. 1470–1635’,

Proceedings of the Royal Irish Academy 84C (1984), 2y115.

46Aonghus Moloney, ‘East to west, crossing the bog at Clonmacnoise’, in H.A. King (ed.),

Clonmacnoise Studies. Vol. 1: Seminar papers 1994 (Dublin, 1998), 2–10: 9; T.C. Breen, ‘Excavations

of a roadway at Bloomhill bog, County Offaly’, Proceedings of the Royal Irish Academy 88C (1988),

321–39.

47Condit and Cooney, Clonmacnoise bridge, w3x.

48I would like to thank Dr Aidan O’Sullivan for discussing dug-out boats found at the

Clonmacnoise bridge with me.

68 Proceedings of the Royal Irish Academy

comparable vessels were recovered at different spots around the lake. Within living

memory, bodies of the dead have been ferried across the Shannon at this point

for burial in Clonmacnoise.49

The first reference to the construction of a barrier or causeway in this region,

as distinct from mentions of already existing structures, is in the Annals of Ulster

for 1001: one was built at Athlone and another at Be´al A´ tha Liag.50 The form of

construction of these bridges is unclear, but they were almost certainly made of

timber. To facilitate their incursions into Meath, Toirrdelbach O´ Conchobair and

his son Ruaidhri built a series of wooden bridges at Athlone and A´ th Liag beginning

in 1120; a number were in turn destroyed by the Ua Ma´el Sechlainn.51 Toirrdelbach

also constructed bridges at Shannonbridge and at Donloe on the Suck.52

In 1129 Toirrdelbach built a castle at Athlone, probably to protect the bridge.53

The first stone bridge at Athlone was that built by the justiciar of Ireland, Bishop

John de Gray, in 1208–9 or 1210.54 This was destroyed by A´ ed O´ Conchobair in

1271 or 1272.55

Fording points and bridges have traditionally facilitated the evolution of settlements;

numerous examples of the phenomenon can be found in Ireland and

elsewhere. Athlone—with a bridge and fortress, of whatever form—is a good example;

Lanesborough is probably another.56 The Anglo-Normans on their arrival built

their castles and mottes at or near crossing places or fords. The control of crossing

places, particularly on navigable rivers, was a strategic goal in the domination of

an area. Cloonburren, near which the Slighe Mho´r route ran east–west across the

Shannon callows, has the remains of the finest motte in County Roscommon,

although references to this structure are from Gaelic Irish sources. Bridges also

have links with religious settlements, whether, as at Abbeyshrule and Cloonowen,

the settlements were sited at river crossings and might have needed a temporary

framework to span the river, or whether, as at Boyle, the bridge was an integral

part of the settlement that evolved at the site.

Roads

In the written sources, such as the Civil Survey, toghers and roads are cited not

as features of interest or importance in themselves but as identifiable landmarks

in the landscape described.57 As well as being necessary for everyday commerce

and living, roads and trackways were strategically important. The documents contain

repeated references to particular routes that were clearly of military significance.

In 1201, when the hosting into Connacht led by John de Courcy and the

49I am grateful to Mr Peter Naughton of Cloonburren for this information.

50Harmon Murtagh, Athlone, Irish Historic Towns Atlas 6 (Dublin, 1994), w1x. Murtagh dates

the first bridge at Athlone to 1120. There are corresponding entries in the Annals of Clonmacnoise,

the Annals of Loch Ce´ and the Annals of the Four Masters. I am indebted to Dr Murtagh for pointing

out the Annals of Ulster reference to me and also for his discussion of the subject.

51ALC, vol. 1, 113; Ann. Clon., 191, 193, 204, 205; Misc. Ir. annals, 13.

52Ann. Tig., 41.

53ALC, vol. 1, 127.

54Ann. Clon., 223; ALC, vol. 1, 245.

55Ann. Clon., 249; ALC, vol. 1, 471.

56SMR RO037–004/005.

57Civil Survey, vol. 10, 49.

DORAN—Medieval Communication Routes 69

son of Hugh de Lacy was defeated by the O´ Conchobair, the Anglo-Normans

retreated by way of Tiaquin in Co. Galway ‘at the head of the Tochar Mona Coinnedha’

and hence to Roscommon.58 This routeway was also used in 1225 by A´ ed

Conchobair as he advanced to meet the Anglo-Normans of Leinster, who were

moving into Connacht in response to his request for assistance.59 In 1255 a great

meeting took place at ‘Tochar Mona Coinnedha’, where a peace was concluded

between the O´ Conchobair and the Mac William Burkes.60 The same routeway

continued to be a meeting place of armies throughout the thirteenth and fourteenth

centuries.61

In his article on roadways O´ Lochlainn listed some devices that may provide

hints to the existence of roads—place and townland names, for example, that

contain either as a prefix or as a suffix elements such as to´char, ‘flagged path’, ceis,

‘wattle path’ or rod, ‘roadway’.62 Rodeen, Co. Roscommon, is an example of the

latter, and bealach, ‘gap’, is recalled in Bealragh in the same county. Topography

would also have influenced the siting of roads: bogs and flooded areas would have

been avoided, but it is common to find a roadway on eskers such as those that

pass through the boglands of Counties Longford and Roscommon. It is important

to remember that unlike, for example, Roman roads, these medieval routes were

not essentially physical entities—thin strips of land with physical boundaries; rather

they were rights of way, sometimes with legal and traditional status.63 Routes

tended to follow the line of least resistance, twisting and turning to avoid poorly

drained areas and land that was easily overlooked. Where there was a hill to climb

or a difficult area to pass through, multiple tracks would develop, the traveller

taking the easiest route. Routes may also have varied seasonally as changing weather

affected the condition of the pathway. Most of the early trackways that survive

represent stretches where the road left the cultivated land and where the evidence

has not in consequence been ploughed out or destroyed (Pls V and VI).

Following the path of modern roads can be misleading: a modern road, built

using contemporary engineering techniques, may run along the bank of a river,

whereas an older road will take the higher ground, where natural drainage is

available. Words such as caol (narrow), coradh (a weir) and eiscir (an esker) in place

names may contain clues to old roads; as do ancient places of assembly and inauguration

that were presumably served by routeways. Cruachain, Co. Roscommon,

is an excellent example of this (Fig. 1).64 Pilgrim roads, such as that to Clonmacnoise

discussed above, are a singular type of routeway formed by a particular kind

of traffic. They may, like the Clonmacnoise road, have had multiple purposes and

58ALC, vol. 1, 221.

59ALC, vol. 1, 285; Ann. Conn., 17.

60ALC, vol. 1, 407; Ann. Conn., 113.

61ALC, vol. 1, 441, 583; Ann. Conn., 139, 243.

62O´ Lochlainn, ‘Roadways in ancient Ireland’, 466.

63Paul Hindle, Medieval roads and tracks (Buckinghamshire, 2002), 6.

64The Cruachain complex is situated on the broad limestone plain known as Mag nAı´, northwest

of the village of Tulsk, Co. Roscommon (Fig. 1). It is mentioned in the early sources as one

of the chief pre-Christian cemeteries of Ireland, as a royal residence and as a place of ritual

assembly. It is the setting for the beginning of the early Irish saga the Ta´ in Bo´ Cu´ ailgne.

Traditionally, it was the inauguration site of the kings of Connacht. The modern field fences

cover the relicts of older field systems and roads around a concentration of important and

unusual earthworks. These remains cover a broad chronological span from the prehistoric

through to the medieval period.

70 Proceedings of the Royal Irish Academy

PL. V—Boher an Choraine, Co. Roscommon.

PL. VI—Trackway at Toberorry, Co. Roscommon.

DORAN—Medieval Communication Routes 71

may have followed the main highway for part of their course. These pilgrim routes

provide important indications of early roadways because they were better maintained

than other tracks that followed corresponding routes.65

When the routeways isolated by O´ Lochlainn that pass through Longford and

Roscommon are plotted, it is immediately obvious that the early ritual site of

Cruachain is a focal point (Fig. 1). Of the sixteen roads taken from O´ Lochlainn’s

work, almost all move towards or radiate from this point. Many are associated with

religious sites: Mocmoyne (Assylin/Essmakirk), for example, is an early and important

foundation. Clergy from Mocmoyne had the right to attend the inauguration

of the kings of Connacht. The correlation between religious settlements and the

roads is notable, but this may be related to the sources used by O´ Lochlainn, the

majority of which are ecclesiastical.66 This region is bisected by two major routeways:

the Slighe Assail and the Slighe Mho´r. The former, which bisected eastern

Roscommon and all of Longford, was a key artery linking this area with the eastcoast

ports and with English and continental markets. The extreme south of County

Roscommon was clipped by the Slighe Mho´r going west to east via Cloonburren.

The two highways are linked by a minor road (Fig. 1, Route 9), as well as by the

River Shannon.

In attempting to followO´ Lochlainn, I have given precedence to his map while

using the locations that he derived from the various sources as a broad guideline

(Fig. 1; see Table 1 for list of roads). Often two or three roads existed between

given points. I have followed as far as possible his outline with respect to the

topography and have used archaeological sites along the routes as markers for

possible directions. Although some of the roads, for example Route 2 from Drumsna

to Mocmoyne, cover long distances, little information about them is provided

in the text of O´ Lochlainn’s article or in his map.67 Here I have taken the profile

of the road on his map and followed the most likely route using the six-inch

Ordnance Survey maps. Where no obvious direction is indicated in his text or

map, as in the case of Route 12, I have followed the path of roads already identified.

Alternative directions exist, of course, and this road could have gone

straight from Elphin to Termonbarry rather than by way of Tulsk. The latter, however,

is an identified route, and it would appear to be the more likely alternative.68

The Slighe Assail was traditionally the main road from Meath to Connacht and

can clearly be seen to be focused on Cruachain (Fig. 1). In County Longford a

number of important Anglo-Norman settlements were established along its passage,

including the borough of Lissardowlan and, on a link road, the borough of

Granard and the Cistercian abbey of Abbeylara. Abbeylara contains one of two

Cistercian abbeys in County Longford. The abbey was founded by Richard de Tuit,

whose motte lies at Granard, shortly before his death in 1210 or 1211; it was

colonised from St Mary’s Abbey in Dublin in 1214. Initially Anglophile in culture—

in fact, the racism of the house towards the Gaelic Irish is cited in the remonstrance

of the Irish princes in 1318—by the early fifteenth century it had passed

65For discussion of the maintenance of pilgrim trails in other parts of Europe, see Jonathan

Sumption, Pilgrimage (London, 1975), 116.

66O´ Lochlainn, ‘Roadways in ancient Ireland’, 467.

67Gwynn and Hadcock give Mocmoyne as Assylin, alternatively as Eas-mac-n-Eirc. The name

in O´ Lochlainn’s text and map for this site may thus be an anglicisation of this name. Aubrey

Gwynn and R.N. Hadcock, Medieval Religious Houses: Ireland (Dublin, 1988), 30.

68O´ Lochlainn, ‘Roadways in ancient Ireland’, 469, section 8 (a).

72 Proceedings of the Royal Irish Academy

TABLE 1—List of roadways shown in Fig. 1.

Highways

Slighe Mho´r Dublin to Clonmacnoise to Ballinasloe; to Cruachain via route 9

Slighe Assail Dublin via Oldbridge to Edgeworthstown, Longford, Cloondara,

Termonbarry, Tulsk and Cruachain

Secondary routes

1 Essmakirk to Ballina (Bo´thar na Sliabe)

2 Drumsna across the River Boyle to Essmakirk and Kesh

(Red Earl’s Road)

3 Elphin to Breckliff via Essmakirk

4 Frenchpark to Coolavin

5 Killarought to Tullinarock

6 Rathcroghan to Elphin, Bellanagare, Frenchpark, Killarought,

Dromod and Ballaghaderreen

7 Rathcroghan to Toomona and Tulsk

8 Rathcroghan to Tuam

9 Ballinasloe to Lough Croan, Roscommon and Boyle

10 Tuam to Roscommon

11 Roscommon to Clonmacnoise, Clonard and Seir

12 Boyle to Longford via Elphin and Tulsk

13 Abbeylara to Granard, Clonbroney, Longford and Cloondara;

Termonbarry to Rathcroghan

14 Uisnech to Ardagh and Longford

15 Nure to Clonmacnoise

16 Slighe Mho´r to east Galway

to the control of theO´ Fearghail and reflected the prevailing political and cultural

environment.69 Abbeylara is sited about 90m west of a stream and overlooks Lough

Kinale; one of the cranno´gs in this lough is close to the find-spot of a decorated

medieval book-shrine.

In Longford there is no record of an Anglo-Irish settlement, but the town was

the site of a later Dominican foundation and a Gaelic market settlement.70 In

addition, a number of mottes and moated sites are found in the vicinity. Although

the Slighe Assail pre-dated the arrival of Anglo-Norman colonists, it was an obvious

bolster to settlement. Because it was a major east–west channel and crossed the

Shannon, it permitted access to the east coast as well as to Athlone and the south.

It is very likely, however, that control of the roadways was as contested as control

of the Shannon. This was always a border area, and the Anglo-Norman settlements

at both Lissardowlan and Granard were short-lived. Nevertheless, when recorded

Anglo-Norman settlement is plotted against the road network, a relationship can

69Gwynn and Hadcock, Medieval religious houses: Ireland, 124.

70Linda Doran, ‘Medieval settlement in Longford and Roscommon’, unpublished PhD thesis,

University College, Dublin, 2001, map 4.1.

DORAN—Medieval Communication Routes 73

clearly be seen (Fig. 2). Lissardowlan was directly on the Slighe Assail, and Granard

was on a side road. No evidence has been found for a link road from the Slighe

Assail to Lanesborough; however, goods produced there could be transported up

the Shannon to the ford at Termonbarry and moved east along the road.

Similarly, the Anglo-Norman borough at Roscommon was at a mid-point on

Route 9 and was connected to both the Slighe Assail and the Slighe Mho´r. Although

the direction of the medieval roads connecting Ballintober to the Slighe Assail or

Route 9 is unknown, they may follow the course of the modern routes. This would

have facilitated the movement of goods and people to the markets and the ports

in the east and south. In County Roscommon, in addition to the crucial focal

point of Cruachain, the Slighe Assail passes close to a number of early Christian

establishments. This area of north Roscommon, which has pockets of good land

embayed in areas of bog, was in Gaelic hands, with no visible Anglo-Norman settlement.

Although the area contains a number of moated sites—normally an indication

of a frontier colony—they are likely to be of native construction.71

In both Longford and Roscommon, minor roads radiate from the Slighe Assail.

In Roscommon a number of these converge on important early Christian and

later ecclesiastical centres: Mocmoyne (just east of Boyle), Cloonshanville, Tibohine

and Elphin, which was the diocesan centre for east Connacht for the greater

part of the medieval period. Mocmoyne is the terminus of a number of minor

secondary roads. One of the most important foundations in the same area was

that of the Premonstratensian house of Holy Trinity in Lough Key. Whereas other

ecclesiastical sites are served by, or are close to, minor routes, some important

early sites, such as Baslick, are situated away from the recorded roads. In addition

to the Cruachain complex, a number of significant Gaelic secular settlements are

concentrated in this section of Roscommon. The island fortress of Carraig Mac

Diarmata in Lough Key, close to the Cistercian abbey of Boyle, was a primary Mac

Diarmata strongpoint. Boyle—called by Stalley ‘the most attractive and stylistically

the most intriguing of all the Irish Cistercian houses’—was a prominent centre

for pilgrimage, an industry no doubt assisted by a good road system.72 There was

also a Mac Diarmata residence on Inishterra Island in the area of these routeways.

The death of a member of the Uı´ Domhnaill family there is recorded in 1343.73

Ruaidhri O´ Conchobair’s house at his longphort of Ardakillin Lough is mentioned

a number of times by the annalists. Six cranno´gs can be identified today on this

partly dried-up lake.

In Roscommon 197 cranno´gs have been recorded by Du´ chas; of these, the vast

majority are north of the Slighe Assail. While the extant remains at Tulsk—a

fifteenth-century castle and a Dominican friary—are late, they are adjacent to

Cruachain, and the remains of an impressive raised rath are situated just southeast

of the castle. The castle was built byO´ Conchobair Rua in 1406, and O’Conor

has suggested that this rath may have been an earlier O´ Conchobair residence.74

In the south of the county, Roscommon town is at the junction of a number of

71For a detailed discussion of this thesis, see K.D. O’Conor, The archaeology of medieval rural

settlement (Dublin, 1998). For literary evidence of Gaelic Irish construction of a possible moated

site, see Katherine Simms, ‘Native sources for Gaelic settlement’, in P.J. Duffy, David Edwards and

Elizabeth FitzPatrick (eds), Gaelic Ireland, c. 1250–1650 (Dublin, 2001), 246–67.

72Roger Stalley, The Cistercian monasteries of Ireland (London, 1987), 87.

73ALC, vol. 1, 643.

74O’Conor, Archaeology of medieval rural settlement, 94.

74 Proceedings of the Royal Irish Academy

FIG. 2—Longford and Roscommon: medieval settlement, c. 1300.

DORAN—Medieval Communication Routes 75

PL. VII—Roscommon Castle.

routes: the main artery going north—Route 9, linking the Slighe Assail and the

Slighe Mho´r—is bisected by Routes 10 and 11 there. Roscommon was an important

early Christian foundation. The Cross of Cong, presented to the monastery by

Toirrdelbach O´ Conchobair in 1132, was probably manufactured at Roscommon

by Ma´elI´su mac Bratdan Uı´ Echan, a member of an ecclesiastical family associated

with Cloncraff in Co. Roscommon.75 The complexity and subtlety of its design

make it one of the last great masterpieces of Irish medieval metalwork. A number

of other pieces are linked to what was clearly an important workshop.76 An O´

Conchobair stronghold may also have been established at Loch na nE´an (Loughnaneane),

a now dried-out lake just west of Roscommon Castle (Pl. VII).77

The Dominican priory of St Mary at Roscommon was founded by Felimidh O´

Conchobair in 1253 and is sited at the crossing point of Routes 9 and 11. Monuments

in the friary include the tomb of Felimidh O´ Conchobair. Only two tombs

of Irish kings survive in Ireland; this is the earlier of the two and the more important.

The most impressive surviving monument in Roscommon is the thirteenthcentury

royal castle. This was an integral part—in conjunction with other royal

castles at Athlone and Rindown and the baronial castle of Ballintober—of the

crown’s strategy to secure Connacht. The Anglo-Norman settlement here was

75P.E. Michelli, ‘The inscriptions on pre-Norman Irish reliquaries’, Proceedings of the Royal Irish

Academy 96C (1996), 1–48: 9.

76Cormac Bourke, ‘The bells of saints Caillı´n and Cuana’, in A.P. Smyth (ed.), Seanchas: studies

in early and medieval archaeology, histor y and literature in honour of Francis J. Byrne (Dublin, 2000),

331–40: 338.

77Ann. Conn., 835.

76 Proceedings of the Royal Irish Academy

established in the late 1290s; it had an uneasy history and was probably abandoned

some time after 1360.78 Whereas the choice of Roscommon as the site for a castle

to inspire awe was most likely dictated by a number of factors, the strategic advantage

offered by these routeways was probably a significant one. It is possible that

a connecting road led to Ballintober; the modern road may follow its route.

After 1250 the crown made a series of land grants in south Roscommon, with

the royal castle at Roscommon as the cornerstone of this policy. Southern Roscommon—

the Fews of Athlone—was part of Geoffrey de Costentin’s grant of 1200;

the area was known as O’Naghtan’s county. Graham has suggested that the motte

at Castle Naghten (Ballycreggan) was constructed to control the route from Athlone

to Rindown, a narrow and arduous way passing through bogland.79 Since

water travel was a safer and more efficient alternative, efforts were made by the

Crown to control access to Lough Ree.80 It is noteworthy that the motte at Castle

Naghten is situated between Routes 9 and 11 running towards Roscommon town.

If its role is linked to the protection of Rindown, then the motte may have been

constructed quite late, since the first documentary reference to a fortification at

Rindown is that to the castle built by Geoffrey de Marisco in 1227.81 Dundonnell,

beside Route 9, has been identified as a possible ringwork by both Barry and

Graham. The impressive earthworks at Rahara, sited midway between the two

roads, are another possible ringwork, although the prospect of an Elizabethan

reuse of these earthwork sites invites caution.82 On Route 11, I have assumed that

the road crossed the Shannon at Athlone; the other options would be Termonbarry

or Lanesborough, and since the next point given is Clonard, this would be

a substantial detour.

In County Longford two minor roads run from the Slighe Assail at Longford

town. Both are linked to key early ecclesiastical establishments: Clonbroney, Granard

and Abbeylara on Route 13 and Ardagh on Route 14. Longford was a principal

seat of the O´ Fearghail, the primary Gaelic Irish family in this area. In 1400 they

founded a friary for the Dominican order; the house became observant in 1429.

During the fifteenth century Longford, along with Granard, developed as a market

settlement in the manner of the earlier Gaelic market at Rockingham. There is

no record of an Anglo-Norman settlement at Longford, although one existed at

Lissardowlan 24km away. The early nunnery at Clonbroney was a significant establishment.

The Life of St Samthann, the foundress of this house, is one of the few

texts dealing with a saint native to the area. Granard was the caput of de Tuit’s

manor, which he held from de Lacy. His large motte and bailey is situated on top

of a hill at the south-eastern end of the town (Pl. VIII). This border motte is the

earliest recorded motte in the region, dating to about 1199. Granard is associated

from the seventh century with St Patrick, who is recorded by Tireachan as having

78ALC, vol. 2, 443, 461; Ann. Conn., 139, 153; T.B. Barry, ‘Late medieval Ireland: the debate

on social and economic transformation, 1350–1550’, in B.J. Graham and L.J. Proudfoot (eds),

An historical geography of Ireland (London, 1993), 99–119: 117.

79B.J. Graham, ‘Medieval settlement in County Roscommon’, Proceedings of the Royal Irish

Academy 88C (1988), 2–38: 25.

80Harbison, ‘Rindown castle: a royal fortress in Co. Roscommon’, 144.

81ALC, vol. 1, 295.

82Graham, ‘Medieval settlement in County Roscommon’, 27–8; Timothy Cronin, ‘The

Elizabethan colony in Co. Roscommon’, in Harman Murtagh (ed.), Irish midland studies: essays in

commemoration of N.W. English (Athlone, 1980), 107–20: 118.

DORAN—Medieval Communication Routes 77

PL. VIII—Granard motte, Co. Longford.

established a church there. Aerial photographs suggest that the motte and bailey

occupy part of a much larger enclosure, perhaps connected with the early Christian

complex at the site. The incorporation of mottes into earlier defended communal

sites is seen elsewhere in Europe in this period.83 The medieval borough

of Granard was situated one kilometre south-west of the modern town, in the

townland of Granardkill.

Tradition links Ardagh with St Patrick—the first bishop, St Mel, was his kinsman.

It was the chief church of the politically important Conmaicne and was

recognised as one of the dioceses of the province of Armagh when Tuam was

separated from Armagh in 1152.84 The diocese of Ardagh, which covers most of

the modern county of Longford, was one of the last in the country to have a

parish network. By the end of the early fourteenth century, the framework was still

not in place. There is no documented road link to the crossing at Be´al A´ tha Liag

at the northern end of Lough Ree. The earliest documentary references to A´ th

Liag occur towards the end of the early historic period, when the site acquired

strategic importance as a gateway between Connacht and Meath. The sources refer

to the building and destruction of bridges across the Shannon at this site by both

the Ua Ma´el Sechlainn and the O´ Conchobair.85 In the early thirteenth century

th Liag was part of the de Lacy lordship of Meath. The strategic importance of

the site prompted the construction of a castle, probably a motte, by Walter de

83Robert Bartlett, The making of Europe: conquest, colonization and cultural change, 950–1350

(London, 1994), 65–7.

84Gwynn and Hadcock, Medieval religious houses: Ireland, 60.

85Ann. Clon., 165, 191.

78 Proceedings of the Royal Irish Academy

Lacy in 1221. In contrast to Athlone, which developed into a substantial settlement

at quite an early date, no evidence exists of any significant settlement at A´ th Liag.

A number of toghers were constructed in the stretch of bog to the east, and it is

possible that one of these was part of the road leading from Ardagh and linking

to the Slighe Assail (Fig. 1).

The relationship between the road system and the settlement distribution is

difficult to measure. The picture presented in Fig. 1 is obviously incomplete, as

the evidence for many early roads has vanished, either because they have fallen

out of use or because they have been obliterated under modern roads. Stout noted

in his study of ringforts that high-status secular dwellings were sited away from the

main routes.86 A similar arrangement is seen in the distribution of tower houses

in these two counties, although some, such as Moydow and Ballinmore, are close

to the junction of a number of roads. It is not possible to know if these roads

were in place in the medieval period. Ecclesiastical and borough or market settlements,

however, are situated close to routeways.87 Both early church sites and the

later order houses follow this pattern. Elphin, for example, is at the convergence

point of a number of roads.

The growth of the Gaelic market settlements at Granard and Longford was

undoubtedly aided by their location (Fig. 1). Longford is at a crossroads on a

main highway, and Granard is on a feeder route. The developments are typical of

the fifteenth-century economic expansion that saw the proliferation of tower

houses and the spread of friaries, particularly in Connacht. Both of these are

indications of disposable wealth created by localised economic and political stability.

88 This economic recovery is particularly notable in areas such as this region

that were outside the ambit of the Dublin government. The success of these market

settlements exemplifies the changes in the Irish economy during this period.

As exports were derived increasingly from pasture rather than tillage farming,

Anglo-Norman towns became more and more dependent on the largely Gaeliccontrolled

hinterland for produce to export. The contact that the regional road

system provided with the markets and port towns of the east, south and west was

essential for this trade.

The proximity of these Gaelic market settlements in Longford to the future

Pale resulted in pressure on the markets of Meath, such as Athboy, Kells, Fore,

Mullingar and Oldcastle. In the late fifteenth century the Irish parliament cited

Granard and Longford by name, along with Cavan, as likely to ‘bring great riches

to the king’s enemies and great poverty to the king’s subjects’. The parliament

forbade English merchants to ‘take any goods or merchandise or to carry any

goods from the said markets, or make any concourse or resort to them’.89 This

legislation provides an insight into how much daily commerce must have taken

place between the two communities in this region of shifting frontiers. This trade

was certainly advanced by the access to these market centres provided by the road

and river system.

86Matthew Stout, The Irish ringfort (Dublin, 1997), 103, fig. 29.

87Doran, ‘Medieval settlement in Longford and Roscommon’, map 7.1.

88Howard B. Clarke, ‘Decolonization and the dynamic of urban decline in Ireland, 1300–

1550’, in T.R. Slater (ed.), Towns in decline, AD 100–1600 (Aldershot, 2000), 157–92: 174–9.

89Stat. Ire., 12–22 Edw. IV, 23.

DORAN—Medieval Communication Routes 79

As well as benefiting these Gaelic market settlements, the communications network

may have facilitated internal commerce. Merchants from port towns would

travel to the Gaelic hinterland, selling their imported merchandise and collecting

local produce to export. A well-known letter addressed to a Galway merchant, John

Blake, who died in 1468, and his wife, Juliane French, asked them to bring their

wine to Roscommon town, where they would be able to sell all their goods and

where the writer had linen cloth he owed them.90 Following the economic upheavals

of the fourteenth century, Ireland, like so many regions in Europe, concentrated

on a specialised trade—in this case cattle, whose hides found a ready

market. Towns such as Roscommon, strategically situated on an access road

between two major highways, were ideally placed to take advantage of these new

economic circumstances.

The road network differs significantly in each county. Longford has a main

road—the Slighe Assail—with two routeways leading from it. (The road network in

Longford is in fact similar today, with one major roadway through the centre of

the county.) Roscommon, on the other hand, has a network of roads; the area is

bisected by the Slighe Assail, and the Slighe Mho´r skirts the southern edge of the

county. In the south of County Longford, a tangle of small toghers run through

the bogs; these, other than the Corlea–Doogarymore road, were mostly for short

internal journeys in the bog and were probably linked to land use. In County

Roscommon, in contrast, an elaborate and complex system of roadways developed,

with a number of important junctions—Tulsk, for example, as well as Boyle, Cruachain

and Roscommon town. The latter, an important site from the early Christian

period, was the terminal point of the Slighe Assail. There were six main

crossing points on the Shannon, in addition to a number of fords. Two of these,

Termonbarry and Cloonburren, had approach roads at right angles on each side

of the river.

Paradoxically for such an apparently isolated region, the juxtaposition of the

Shannon with the two major routeways—the Slighe Assail and the Slighe Mho´r—

meant that it had relatively easy access to the outside world throughout the medieval

period. The Slighe Assail, which linked this part of Connacht with the east-coast

ports and with English and continental markets, was a vital artery. The Slighe Mho´r,

although it only clips the southern end of County Roscommon, was the primary

routeway across Ireland. In addition, this road could be accessed from other parts

of the area: Route 9 links directly to it, but it can also be accessed by sailing down

the Shannon and joining the road at the Cloonburren bridge. This combination

of land and water travel may have been a less arduous and hazardous alternative

than tracking across the bogs. The Shannon is a critical factor to this communications

infrastructure and therefore to the settlement of the area. Throughout

recorded history, the Shannon has been a vital conduit, allowing access to the

outside world and facilitating trade and transport. Without this most dominant

feature in the landscape, and without the juxtaposition of two major roadways

with their feeder routes, the settlement pattern of the surrounding area would

undoubtedly have been radically different.

90K.W. Nicholls, ‘Gaelic society and economy in the high middle ages’, in T.W. Moody et al.

(eds), A new histor y of Ireland. Vol. 2: Medieval Ireland, 1169–1534 (Oxford, 1987), 397–435: 419.

80 Proceedings of the Royal Irish Academy

ACKNOWLEDGEMENTS

This paper is based on work undertaken as part of a doctoral thesis; I would like

to thank my supervisor, Professor H.B. Clarke, for his suggestion that I include

communication routes as an element of that study. I am also indebted to him for

his insightful comments on this paper. I would like to thank Professor T.B. Barry

for proposing the topic as the subject of an Academy paper. I am grateful to Mr

Bill Doran for technical advice on the maps that accompany this paper and to

Billy Doran for advice on the computerisation of the data that form the basis of

the study. I am indebted to Ms Louise Doran for the photographs. The people in

both counties who showed us kindness and cooperation are too numerous to list

here. I would like, therefore, to thank them collectively, in particular those who

took the time to walk with us and share their knowledge of the landscape. I am

indebted to Mrs Rose Forde, who not only encouraged me to undertake the study,

but also provided the support needed to accomplish it. My thanks are also due to

Roscommon County Council and Longford County Council for financial assistance

with the publishing costs of the colour maps produced as part of this paper.

ABBREVIATIONS

ALC W.M. Hennessy (ed.), The Annals of Loch Ce´: a chronicle of Irish affairs

from A.D. 1014 to A.D. 1590 (2 vols, London, 1871).

Ann. Clon. Denis Murphy (ed.), The annals of Clonmacnoise, being annals of Ireland

from the earliest period to A.D. 1408, translated into English A.D. 1627 by

Conell Mageoghagan (Dublin, 1896; facsimile reprint, Lampeter, Wales,

1993).

Ann. Conn. A.M. Freeman (ed.), Anna´ la Connacht: the annals of Connacht, A.D

1224–1544 (Dublin, 1970).

Ann. Tig. Whitley Stokes (ed.), The annals of Tigernach. Facsimile reprint from

Re´vue Celtique 16–18 (1895–7).

Cal. justic. rolls Ire. Calendar of the justiciar y rolls, or, proceedings in the court of the justicular of

Ireland, 1308–14 (Dublin, 1956).

Civil Survey R.C. Simmington (ed.), The Civil Survey, A.D. 1654–56 (10 vols, Dublin,

1931–61).

Irish exchequer payments Philomena Connolly (ed.), Irish exchequer payments, 1270– 1446 (Dublin,

1998).

Misc. Ir. annals Se´amus O´ hInnse (ed.), Miscellaneous Irish annals, A.D. 1114–1437

(Dublin, 1947).

Stat. Ire., 12–22 Edw. IV James F. Morrissey (ed.), Statute rolls of the parliament of Ireland, twelfth

and thirteenth to the twenty-first and twenty-second years of the reign of

King Edward the Fourth (Dublin, 1939).

*Present address: 7 St Mary’s Road, Ballsbridge, Dublin 4.

E-mail: linda@billdoran.net

Proceedings of the Royal Irish Academy, Vol. 104C, No. 3, 57–80 (2004)  Royal Irish Academy