Chap. 2: History and Geography of the British Isles to the End of the Roman Period
The oldest name under which Classical writers commonly refer to the British Isles is The Pretanic Islands. They are thus referred to by Pytheas in the late fourth century BC, and by later writers – Polybius, Strabo, Avienus. The form implies for the name of the inhabitants Pritani or Priteni. The name is probably Celtic and of the Brythonic form – P-Celtic – probably Gaulish; the Gauls may have handed this name on to the Greeks. The older form continued in use in Welsh texts referring to the island as a whole, Prydain, Britain from Pritani; and a variant from Priteni is used in the form Prydyn in early Welsh texts referring to the people north of the Antonine Wall, the Picts. I In the Roman period the people of the Roman province called themselves Brittones, perhaps a corruption of Pritani .
The peoples who brought the Celtic languages into the British Isles first come before us as living inhabitants in the pages and atlas of  Ptolemy, a Greek geographer of Alexandria who flourished about the middle of the second century AD, and whose geography, including his account of Ireland, is based on the lost work of Marinus of Tyre, who lived earlier in the same century, though the information which Ptolemy gives about Ireland is believed to be derived ultimately from the work of some traveller of at least two centuries before his time. The information of Marinus is probably that of the first century AD and perhaps earlier; but various strata of information can be detected, including some later details. It may be added here that in addition to Ptolemy we have a valuable source for the geography of Celtic Britain in an anonymous work of the sixth century AD known as the Ravenna Cosmography, which makes use of sources of Ptolemys time, and of various other periods.
From the nature of Ptolemys record we gather that, as in pre-Roman Gaul, the Celtic peoples in the British Isles were divided into a number of separate groups, of whom the fifth century geographer Marcian states that Ptolemy enumerates thirty-three in Britain, seventeen of whom belong to southern Britain. Scotlande therefore probably had about sixteen. Ptolemy gives the names of nine cities (tribes or states) with their relative positions on the Irish coast, but the latter are of uncertain authenticity, and are presumably derived from sailors records. He names none away from the coast, and all the names which have been identified belong to the east and south coasts, while for the West coast too little information is given to supply us with even an impression of its occupants. The northern and west coast of Ireland is one of the stormiest in the world, and it was probably avoided as far as possible by sailors. The cities which Ptolemy names were probably places of assembly or royal raths – oppida, like Tara and Emain Macha, and two are referred to as regia . None can be identified with any certainty, and other names suggest close relationship with corresponding tribal names in Britain. The Brigantes in South Wexford can hardly be dissociated from the Brigantes who occupied most of north Britain in the Roman period. In general, the names in Ptolemys account leave no  doubt that Ireland was already Celtic-speaking when they were recorded. Very few of Ptolemys tribal or city names can be identified with names in Irish literary tradition, and therefore his Ireland is earlier than that of the Heroic Age. His Irish names have been derived directly or indirectly from some geographer considerably before Ptolemys time, and Pytheas is the only competent candidate.
We have no records of any languages earlier than Pictish in these islands. All our evidence – linguistic, archaeological and historical –suggests that in pre- and proto-historic times the Picts were a great nation occupying the northern part of Scotland from the Firth of Tay to the Shetlands, whom Bede knew in the eighth century AD as the Northern and Southern Picts. Their origin is unknown, but they had probably had a continuous history in Scotland from the Bronze Age until the movements of the Celts into Britain, with whom they seem to have combined. Their proper names have been preserved on about two dozen stone inscriptions evenly distributed between the Northern Isles and the Tay, almost all written in a late form of the early Celtic alphabet known as ogam . Apart from the proper names the inscriptions have never been interpreted. The evidence of the proper names, and the few Pictish words which have survived, suggest that the Celtic element in the Pictish language was nearer to Brythonic than to Goidelic, and nearer to Gaulish than to Welsh. This Pictish language continued to be spoken as late as the ninth century, till the end of the Pictish hegemony.
It may be conjectured that in the British Isles, as in Gaul, some of the earlier Celtic kingdoms entered on a large scale, and separated into detached groups either before or after establishing themselves here. This is possibly the explanation of the curiously wide distribution of some early identical names. Thus we find the Cornavii recorded by Ptolemy in the extreme north of Scotland, in the area of modern Sutherland or Caithness, and the Cornovii on the north Welsh border with its oppida recorded by Ptolemy as Deva (Chester) and Viroconium (Wroxeter). The first element Corn- appears in other places in western Britain, including of course Cornwall, as well as Cornouaille in Brittany. Another ancient name of wide distribution is that of the Dumnonii located by Ptolemy in the west of Scotland south of the Forth-Clyde line, but reaching into Stirlingshire. He also assigns the whole of the southwestern peninsula of England, however, including the modern Cornwall, Devon, and much of Somerset, to the Dumnonii. In Ireland a people in the ancient kingdom of Connacht and part of Leinster were known as the Fir (Men) Domnann, but not by Ptolemy. Are these the relics of early Atlantic settlements? Ptolemy marks three other tribes in southern Scotland, the Novantae in Galloway, south of the Damnonii; the Otadinoi in Haddington and Berwickshire and south into Northumberland; and the Selgovae between the two in the hill country forming the watershed between the Clyde and the Tweed.
In Wales our information of the tribal divisions before the Norman Conquest is very incomplete. The principal tribes located by Ptolemy lie along the sea-coast, and no doubt include the mountain zones in the background which receive no separate names. From Ptolemy and other writers we learn that the south-east down to the Bristol Channel was occupied by the Silures. Their tribal centre was evidently at Caerwent (Venta Silurium). Ptolemy refers to the Demetae (later Dyfed) as in the extreme west. For north-western Wales our knowledge is not very precise. The whole of north Wales may lie within the territory of the Ordovices, and Ptolemys information for Anglesey and Snowdonia is probably defective. To the north-east were the Decangli of Flintshire, referred to by Tacitus as dwelling not far from the sea which looks towards Ireland. 
Caesars initial expedition to Britain in 55 BC has as its object an appraisement of the situation. He himself tells us (Book IV, 20) he understood that in almost all the Gallic campaigns help had been forthcoming to the enemy from that quarter, and he believed it would be of great advantage to him to have entered the island and observed the character of the natives, and the nature of the country which was little known to the Gauls. He tells us that as soon as the  Britons observed his intention of landing they sent forward their cavalry and charioteers, whom it is their regular custom to employ in fights, and followed them up with the rest of their forces. The chariot fighting of the Britons evidently made a deep impression on Caesar, for after relating a difficult landing and a subsequent defeat of the British forces, he pauses in his narrative to give a somewhat detailed account of the method of their chariot tactics:
Caesar had made a reconnaissance in force in his two expeditions to Britain in 55 and 54 BC. The actual conquest began under Claudius in AD 43, and spread northwards and westwards till the Celtic tribes of the English lowlands seem to have been over-run and subdued as far as the Severn and the Humber within three years. In 47 the Iceni were conquered, and for the most part the older established tribes, including the Iceni themselves, accepted the position of client kingdoms. In the north also Cartimandua, the queen of the Brigantes, entered into treaty relations with Rome, and accepted the position of a client state.
The ancient Celtic kingdoms beyond the periphery of Lower Britain, however, were in no sense reconciled to a conquest by the Romans. Resistance movements and rebellions broke out whenever and wherever opportunity seemed to offer any chance of success. To follow the history of the Celtic kingdoms as the Roman conquest spread is particularly difficult as our only records are those of the conquerors, and we are in the position of seeking to trace an almost universal underground movement for which we have no continuous clues. It is as if one should seek to trace the course of a series of small earthquakes without the guidance of seismological instruments. We are thankful for such surface records as Roman historians have left us.
[Account ensues of campaigns in Wales and Scotland; the resistance of Caratacus, who was handed over in chains to the Romans by Cartimandua, a women-ruler and contemporary of Boudicca; the conquest of Britain by Agricola in 77 and 78. Dio Cassius portrait of Boudicca, called the most dramatic picture of a Celtic heroine in classical literature (G. R. Dudley & G. Webster, The Rebellion of Boudiccr, 1962)]
[Quotes Tacituss account of the Ordovices defeat by Suetonius Paulinus on Anglesey, where women and druids played a part in the Celtic battle-array:]
[Also Dio Cassiuss description of Boudicca:]
[Of Agricola:] Reluctantly as he looked at the opposite coast of Ireland, clearly visible, he abandoned, or perhaps rather postponed, the dream of conquest yet further westward. As Tacitus, who was in a position to know, tells us, He saw that Ireland … conveniently situationed for the ports of Gaul, might prove a valuable acquisition, and Tacitus goes on  to tell us that one of the Irish knigns, who had been forced to flee owing to local political troubles, was received by the Roman general, and, under the pretence of friendship, detained to be of use on some future occasion (Tacitus, Agricola, XXIV).
‘I have often heard Agricola declare, adds Tacitus, that a single legion with a moderate band of auxiliaries, would be enough to complete the conquest of Ireland. (Ibid.) 
[Gives account of the conquest of Scotland, the shifting border and barriers - walls and ditches - made; battle of Mons Graupius, 84 AD; Hadrians Wall; pax Romana; weakening of Roman hold on Britain; Picts and Irish gaining strength; both possessed fleets more numerous and powerful than is commonly realized, and both were attacking British coasta throughout this period [31-32]; de curicus (currachs)]
[Quotes Calgacus's harangue to his army facing the Romans at Mons Graupius, 84 AD, as reported in Tacitus:
[Fall of Britain; Constantine usurps title of Emperor; crosses into Gaul by friendly arrangement with Emperor Honorius; defeated and slain with his sons; rescript of Honorius, 410 AD, informing cities of Britain that they might look after themselves, and allowing natives to bear arms .]
In consequence of her seclusion and inviolability Ireland has preserved in miniature for future ages the economy and institutions of the ancient Celtic world of the Iron Atge. She is the microcosm of the early Celtic race. The art of La Tène continued without a break to form an organic element in the Irish Art of the Middle Ages. The intellectual classes of Gaul reappear in ancient Ireland to carry on the intellectual life and historical traditions by oral transmission till written records begin with the introduction of Christianity and Latin learning. In the absence of Roman penetration of Ireland her political geography changed only slowly, and we lack the external evidence of either contemporary Latin documents or early vernacular written records to help us to trace the early settlement and population groups.
On the other hand Ireland possessed a greater wealth of carefully preserved oral tradition from the earliest period of our era than any other people in Europe north of the Alps. For this reason the foundation of her early history from traditional materials is of general interest far beyond her geographical and political area, and second only to that of the ancient Greek and Roman world. Further, archaeologically speaking Ireland is a museum of the ancient world. The Industrial Revolution which changed the face of Britain left the Irish countryside inviolate, a paradise for the archaeologist and place-name specialist. The absence of Roman disruption and of recent industrial development combine with the discipline of modern scholarship to enable us to reconstruct conjecturally something of the early settlement of Ireland before the days of written history. The process of reconstruction is under way, and it may well add centuries to the history of the periphery of western Europe.
When the Irish historical period begins early in the fifth century the dominant peoples of Ireland were the so-called Goidels, with strongholds at Tara in Meath, Croghan in Connacht, and Cashel in Munster. The chief Irish families claimed to be of Goidelic origin, and to be related to one another closely through the parent stock. They were a group of leading families, over-kings, who held their lands free of rent or tribute, while themselves exacting tribute from the provincial subject kingdoms.
Our oldest genuine traditions, however, relate to a period which has been credibly shown to be that of Ireland during the previous period, and in fact to be, not learned speculation, but genuine oral tradition, embodying a reliable picture – within certain well-defined limits – of Ireland as she was before the introduction of writing, of European culture, and before the dominance of the dynasty of Meath over the north. The historical geography and political institutions of this traditional picture of Ireland offer a remarkable contrast, on the one hand to the picture artificially built up by the late pseudo-historians, working in the interest of the historical dynasty of Tara, and an equally strong contrast on the other to those conditions known from genuine historical records, to have prevailed from the fifth century onwards, as we shall see. 
This traditional picture is that of a heroic society in the late Iron Age, much like that of Gaul before the Roman conquest. It has been preserved for us intact by the exceptionally high quality of Irish oral tradition. This phase of Ireland which immediately preceded the historical period merits our keen interest and attention. It offers us a picture of Celtic society which is unique in Europe.
This earlier oral tradition claims to extend reliable Irish history back to the Heroic Age, and it can be shown to be appropriate to a period as early as the fourth century AD. Many traditions may even be authentic for the close of the third century, for example the person and some of the stories about Cormac mac Airt, who according to tradition reigned as king of Ireland from 227 to 266. Our fullest and most authentic picture of Ireland in the earliest historical period, however, is the great prose saga of Táin Bó Cualnge, “The Cattle-Raid of Cooley”, which has preserved for us an intimate record of a European society in the late Iron Age. Ireland is already a Celtic country in this story, as the proper names prove. We must realize, however, that the story has come to us through various centuries of oral transmission, during which the earliest tales, originally formulated in the Heroic Age, had been recast in a final artistic form, first oral and finally, probably in the early eighth century, in written form. The storyteller recognizes the period as belonging to the far past, but the allusions and traditions have been preserved with great fidelity, as can be seen from the consistency of the internal evidence, and from comparison with other Irish heroic stories.
In the Táin Ireland is divided into four provinces, already called cóiceda, fifth, each province consisting of one fifth; but the tradition is so conservative that the fifth province, known in historical times as Meath (Mide), and which seems not yet to have come into existence, is never mentioned. The Ireland of the Táin is the Ireland of the Erainn, the earlier Celtic inhabitants, and it is thought that the fifth province, Meath (Ir. Mide, the Middle Province) was added later after the establishment of the supremacy of the Uí Neill (‘descendants of Niall Noigiallach) in central and northern Ireland in the fifth century. The four provinces known to the Táin tradition were large kingdoms, like the oldest Celtic provinces of Gaul and Britain, and they bore the names Ulaid, Connachta, Laigin and Mumu. The names of the four-fold division have survived till the present day as Ulster, Connacht, Leinster and Munster. The province of the Ulaid (‘people of Ulster) consisted In the Táin of the whole of northern Ireland, including Donegal in the  far west and Dunseverick on the Antrim Coast. It was a great monarchy with its chief ráth or court at Emain Macha, tow miles west of Armagh, and its king was Conchobar mac Nessa, a prince of the Érainn.
The second of the Cóiceda at this period was Connachta (modern Connaught), with its centre of power spreading widely over the hill of Cruachan. In the Táin Connachta is ruled by a woman, Queen Medb, whose consort is Ailill, and the province of Connachta is the rival and enemy of the Ulaid. South of the Ulaid were the Laigin (modern Leinster). The Cóiced of Mumu (modern Munster) is ruled by small kings of whom the chief is Cú Roí mac Dáiri, whose seat is in West Kerry, and who has close relations with the heroes of the Ulster Cycle. As yet the Eoganacht dynasty of the great rock citadel of Cashel has not arisen.
The tradition represented by the Táin and its framework of the four great provinces gives us our first introduction to early Irish history, comparable to our picture of Britain at the time of the Roman conquest. Christianity is unknown. Tara plays no part. The Goidels are never referred to. Everything in the Táin suggests that the political and religious conditions preserve an unbroken tradition from before the fourth century, a Heroic Age civilization of the sub La Téne period. For these reasons the oral traditions preserved in the Táin may be said to represent the earliest period of Irish history, which dates from the fourth century while the Heroic Age was still a living memory.
Moreover, the sympathies implied in the Táin are all with the Ulaid and therefore with the Érainn, not with the later dynasty of Tara. The Ulster places and the Ulster heroes are more familiar to the narrator than those of Munster and Connaught. This suggests that the Táin is an Ulster composition, and as the monastery of Bangor, Co. Down, was a centre of historical studies in the seventh and eighth centuries, it is probable that the Táin may have been finally written down in this monastery. And the motive for the compilation of this great prose epic? Perhaps we may suggest that it is a proud assertion of the past greatness of the Érainn, the ancient Celtic rulers of Ireland, against the parvenu line of the Ui Neill of the Tara dynasty, and the northern Ui Neill, who had destroyed Emain Macha and reduced the Ulaid to a small territory in northern and eastern Antrim. It may be suggested that the final formulation of the story and its record on vellum was inspired by a political motive.
Subsequently to the period represented by the Táin, a new dynasty arose with its royal ráth at the prehistoric sanctuary of Tara in Meath. It was apparently the founding of this kingdom which established the division into five fifths (cóiceda). This was indeed of short duration. The founder of the new dynasty was Mall Noigiallach, and his descendants  subsequently ruled all the centre and northern half of Ireland. The southern half, consisting of half Leinster and all Munster, was never subject to the Uí Néill (descendants of Níall), but at a later date was ruled by a branch of the Munster dynasty from the great rock of Cashel. It is believed that the new dynasty of the Uí Néill and the province of Meath rose to prominence as the result of invasion, and that they were Goidels. Who were the Goidels?
The ancestry of these Meath Goidels is traced in late pseudo-historical texts to a certain fictitious Míl Espáine who was supposed to have come from Spain. Hence the name Milesians, by which name the
Goidelic conquerors are commonly referred to in modern histories. Míl Espáine is evidently a translation from Latin miles Hispaniae the soldier from Spain, and in the Book of Invasions he is given a fictitious pedigree going back to an eponymous ancestor, Goídel Glas . In fact the name Goídel is borrowed from Welsh Gwyddel Irishman. It may be derived from the Welsh gwydd wood, forest.
The late Professor ORahilly categorically stated that the people of this last invasion of Ireland and the Goidelic form of the Celtic language were introduced simultaneously into Ireland at a comparatively late date, and he uses the term Goidelic of both the people and their language. He expressed the view that: If anything is certain about the Goidels, it is that they reached Ireland direct from the Continent, and he held that they must have come to Ireland from Gaul, and suggested that their migration was from south-eastern Gaul, perhaps from Gallia Narbonensis, whence they passed to the western coast not later than 120 BC, and migrated to Ireland towards 50 BC. (Early Irish History & Mythology, pp.2007-08.) The late date and the course of events suggested by ORahilly are difficult to accept in view of the silence of Roman historians on so long and important a migration at this late date. The question must be regarded as at present unsettled. A more acceptable theory would be that the Milesians (Goidels) did come from Spain, as the Book of Invasions says. The discovery of Q-Celtic inscriptions in Spain by Tovar shows that in that lateral area the older forms had survived.
If, however, we accept the provisional date suggested in Chapter I (p.4f.), that is to say some time early in the second millennium BC, for the spread of the Celtic peoples to the British Isles, it becomes reasonable also to accept the conclusion that the differentiation of the Celtic languages into two principal groups, known as Goidelic (commonly called Q-Celtic) and Brythonic (commonly called P-Celtic, cf. p. 206) originally took place, not, as formerly believed, exclusively somewhere on the Continent, but also in the British Isles at some period  between c.2000 and 600 BC. Very possibly analogous changes may have been taking place throughout the whole Celtic world, and the Quariates may have represented a backward mountain dialect. But, language apart, who are the peoples commonly known as the Goidels of Meath and commonly regarded as the latest immigrants into Ireland large enough to forma distinct population group?
At present this question cannot be answered with confidence; but the following facts are significant:
During the fifth century and earlier the weakening power of the Roman Empire gave a great impetus to the movements and expansion of the Insular Celtic peoples, always seeking to extend their territory. The Irish in particular took advantage of the military weakness of the Roman defences of western Britain to penetrate the western promontories of Britain. We have seen from the statement of Ammianus Marcellinus that already during the fourth century Irish raids were no less formidable than those of the Picts and Saxons from the north and east. The Irish Sea had always been, and continued to be, a purely Celtic area, both politically and culturally. Here the Romans had no  part, and there was no unified Celtic power to preclude the free transit and transport of the individual Celtic communities to and fro across this Celtic Pond.
Among the most significant of these Irish expansions was the occupation of south-western Wales, the area of a wider Pembrokeshire, later known as the kingdom of Dyfed. This area was occupied by a dynasty from Leinster, who led a migration and settled in sufficient numbers to render the population bi-lingual in the fifth century, and to form what was virtually a little Irish kingdom. The pedigree and the story of its foundation are exceptionally well preserved, and these and other traditions enable us to trace also Irish settlements from the same quarter all the way along the south coast of Wales and the Bristol Channel, and even into the little mountain kingdom of Brecknock in the interior. At the same period Irish penetration of the Caernarvonshire peninsula has left abundant traces in place-names, and in the archaeological evidence of Roman defences. We have already traced some of these late Roman defensive measures in the preceding pages.
The peninsula where this Irish influence is strongest is Dyfed. […; 39.]
By far the most important of the Irish expansions of the fifth century was that from north-eastern Ireland to the islands and coast of the south and south-west of Scotland. The coast of either country is clearly visible from the other, and we have seen in Tacitus how the Roman general Agricola had looked across from the Ayrshire coast and speculated on the ease with which Ireland could be conquered. On the other hand as we trace the Irish settlements up the west coast of Britain, and the defensive measures adopted by the Romans against Irish penetration, the expansion of the Irish from the Antrim coast into western Scotland in the fifth century becomes a natural sequel. Irish consolidation on a wide scale in western Britain was only possible beyond the Roman frontier of the Antonine Wall. Here a permanent I rish kingdom was founded in the fifth century, but whether by military  or peaceful penetration is uncertain, though no tradition of military conquest has come down to us. The expansion and consolidation of this Irish settlment and Irish dynasty as the Kingdom of Dál Riata will be traced in the following chapters.
We have seen that according to Zozimus the weakening power of the Roman Empire had stimulated the people of Armorica to follow the example of the Insular Britons and throw off the Roman yoke. During this period, probably already in the fourth century or earlier, the colonization of western Armorica from the south-west of Britain began to take place, and continued without opposition form the Romans till the British language superseded that of the original Gallo-Roman peoples of Armorica. The British emigration has commonly been attributed, on t he testimony of the sixth century British historian Gildas, to the settlement and establishment of the West Saxons in eastern England. A closeer study of the pressure and penetration of the Irish on the western peninsulas of Britain, however, leaves no room to doubt that the British emigration was yet another result of the Irish pressure on western Britain in this period. […; 42]
[End Chap. 2.]
Note: ORahilly points out that the name Mide is late. The old name of the central province is uncertain, Early Irish History & Mythology, p.174, n.4. For the Five Provinces see also ORahilly, Celtica I, p.387f.