The attempt to present the Celts in history as one people, with a common tradition and a common character, is new, and in some degree, experimental. It seems to us to have been justified beyond our expectations, inasmuch as there does emerge in the history and institutions and religion, in the art and literature, perhaps even in the language, a quality that is distinctive and common to the Celts of Gaul, of Britain and of Ireland. We hesitate to give it a name: it makes a contrast with Greek temperance, it is marked by extremes of luxury and asceticism, of exultation and despair, by lack of discipline and of the gift for organising secular affairs, by delight in natural beauty and in tales of mystery and imagination, by an artistic sense that prefers decoration and pattern to mere representation. Matthew Arnold called it the Celtic Magic. (Pref. [ii-iii].)Chap. 1: Discovery of the Celts
Herodotus, in the fifth century, twice mentions the Celts (the first occrrence of the name Keltoí), only to say that the Danube has its source - which he seems to suppose is in the Pyrenees - among the Celts, and that they dwell beyond the pillars of Hercules, and are the most westerly people in Europe except the Cynesians. 
Over all this area, Austria, Bohemia, southern and western Germany, and France, place-names occur which are compounded of such elements as briga - dunum - magus - nemeton - ritum - seno - uindo [list and vars. abbrev.] 
[…] appearance of urnfields in south central Europe … a cultural continuum from the time when they appear, to the close of the Bronze Age, through the Hallstadt period and down to La Tène. But this is too late, in our opinioin, for the separation of Celtic as a distinct dialect, and it is better to regards as already Celtic whatever can be dated to the beginning of the second millenium BC. 
Quotes Tacitus: ‘their Gaulish speech proves that the Cotinin were not Germans. (Germania, p.43.) 
The Celts then are a people who appear in history in the sixth century BC. The early writers mention them only casually, as when Xenophon writes of Celts who fought as mercenaries against the Thebans in the Peloponnesus in 369 BC, and Plato includes them in a list of barbarian peoples who are given to drunkenness. Aristotle says: it is not bravery to withstand danger through recklessness, as when the Celts thake up arms to attack the waves. (Nicomachean Ethics, III, 7 §7; but variously quoted infra.) 
With regard to the Celtic settlement of Ireland in particular, ORahilly in his Goidels and Their Predecessors, and again in his great book, Early Irish History and Mythology, put forward an entirely new theory. He distinguished four successive immigrations: the Cruthin some time before 500 BC; the Cruthin some time before 500 BC; the Érainn (= Fir Bolg) perhaps in the fifth century; the Laigin (with Domnainn and Gálioin) in the third century; the Goidil who came c. 100 BC. The notion of a series of invasions is traditional, and is first recorded by Nennius who knew of three, those of Partholón, Nemed, and the Children of Míl.
In the Book of Invasions as we know it from the eleventh century century recension there are five (or six): (Cesair), Partholón, Nemed, Fir Bolg, Tuatha Dé Danann, and the Children of Mil. But these legends are mere learned fiction, and have no value as history. Their interest is of another kind. The only bearing they have upon history is in so far as they betray a state of affairs which the literary men were trying to expound. What does appear from various references cited by ORahilly is the presence in Ireland of at least three ethnic groups: Cruthin (or Cruithni) who were mainly in the north-east and plainly akin to the Picts of Scotland (also called Cruthin); Érainn, who were mainly in the south-west and in the south-east corner (Déisi); and Goídil (Cland Míled) who reigned over Tara, Cashel and Croghan and were the dominant people in the early historic period. Whether the Laigin were a distinct people or merely a Goidelic tribe is not clear to us.
ORahillys most novel suggestion is that his first three groups spoke Brythonic dialects, that is that they were P-Celts, and that the Goidil were the only Goidelic speakers, the only Q-Celts. And he seeks to prove this from the evidence of Irish words that are not Goidelic in form and may be Brythonic. His demonstration is not convincing, and the notion that this more archaic language was brought latest, by a migration of the Quariates from south-east Gaul, is inherently improbable. We think it more likely that Goidelic was first established in Ireland, and that Brythonic tribes made settlements there, just as Irish settlements were made in Wales both north and south. In those early centuries, when there was constant intercourse across the Irish Sea, there  was every chance of linguistic borrowing. […] ORahillys doctrine has been accepted by some scholars (M. A. OBrien, Early Irish Society, p.37) and dismissed by others (Vendryes, Etudes Celtiques, I, 352ff.) 
[…] The shrine at Delphos was plundered in 278 […] At that time a territory stretching from Ireland to Galatia was in Celtic hands. ‘For two centuries, says Grenier, ‘they (the Celtcs) were the greatest people in Europe … About 300 BC the power of the Celts was at its height and seems inexhaustible in energy and in manpower. 
This rapid expansion over an enormous area implies great fecundity and a great spirit of adventure. Moreover, like the Greeks in the Mediterranean, the Celts brought with them their civilization, and they imposed it upon the lands they occupied. And though they were barbarians in the strict sense, theirs was no mean way of life. We have accounts of it from Polybius in the third century BC, from fragments of Posidonius who lived in the first century BC, and from Julius Caesar. Some of what Posidonius tells us is worth quoting at length, because it finds such extraordinary confirmation in later Irish sources.
Polybius tells of the conquest of Cisalpine Gaul with a good deal of detail, and of the later struggle of the Romans with the Celtiberi in Spain, and the defeat of the Galatians in Asia Minor. He describes the weapons of the Celts, their preference for fighting naked, the habit of  taking the head of a slain enemy, and he mentions the splendid gold torques and bracelets worn by Celtic warriors. He also says that they attached great importance to clientship as the measure of a mans rank in society.
The best known passage describing the Celts is that in the sixth book of Caesars Gallic War, but Posidonius, who wrote a continuation of the Histories of Polybius, is the main source of information. His works are lost, but Diodorus Siculus and Strabo have preserved a good deal of what he said about the Celts. Here is Strabos account of the Celts of Gaul:
Diodorus describes the habits of the Celts of Gaul at a feast:
Athenaeus, who names Posidonius as his authority, confirms the statement about the Heros Portion:
This practice is vividly recorded in the Old Irish saga of Mac Da Thós Pig (inf. p. 249), and it is the principal theme also of Bricrius Feast.
Another passage is of special interest from this point of view. It is the account by Diodorus of the Gauls in battle:
Here again the Irish sources match closely the description of the Celts of Gaul. The warriors of the Ulster sagas go into battle in chariots, and the challenge to single combat is the central theme in the great prose epic Táin Bó Cualnge. Cú Chulainn, with Loeg his charioteer, in his two-horse chariot, armed with shield and spear and sword, is a Celtic warrior just such as Posidonius describes.
And here is a description of the Gauls by Diodorus, also based, it seems, on Posidonius:
The testimony of Julius Caesar, which is widely familiar, is the most valuable in some respects. He presents Gaulish society as divided into three classes, druides, equites and plebs, the three functions (priest, warrior and husbandman) of which Dumézil has made so much. And of the equites he says that the nobler and wealthier have numerous clients about them.
Caesar tells more about the druids than do the others, their twenty years of study, their oral tradition, and so on. And he says of the Gauls: Natio est omnis Gallorum admodum dedita religionibus ‘The whole Gaulish people is much given to religion. We are told that the druids taught the doctrine of transmigration of souls, that they were concerned with the worship of the gods, and that young men flocked to them for training.
This class of professional learned men, priests and scholars, seems to share a common Indo-European inheritance with the brahmins of India, for the later and fuller evidence of Irish sources shows a similar class, the filid, who, while shorn of their priestly office in a Christian society, have retained the scholarly functions of the druids as poets, genealogists, lawyers, and the practice of oral rather than written tradition. Both the form and content of their learning show astonishing similarity to brahminical tradition.
The archaism of this tradition can best be shown from the Irish evidence which is presented later (chap. 10), but two matters deserve mention here as requiring the assumption of a learned tradition inherited by the druids of Gaul and transmitted by them to the filid of Ireland.
The metres of the Rigveda, the earliest known forms of Indo-European verse, are based upon a line with a fixed number of syllables, of which the first half was free and the cadence was fixed in the form [symbols for short short long SHORT] It was shown long ago by Meillet that the Greek metres have the same origin, and more recently Roman Jakobson has traced it in modern Slavonic verse. Calvert Watkins has now demonstrated, we think convincingly, that the Old Irish heptasyllabic line derives from this Indo-European form, and that other Irish metres are variants of it, thus confirming the antiquity of Celtic tradition, and the common heritage of the druid and the brahmin. (‘Indo-European Metics and Archaic Irish Verse, in Celtica VI, p.194.) 
Dr. Binchy has drawn attention to the resemblances in certain points between the Irish and Indian law-books. In both countries the law consists of canonical texts, invested with a sacred origin, and interpreted exclusively by a privileged caste. There were law-schools in each with varying traditions of interpretation. The relations between pupil and teacher (Irish felmac and fithithir; Sanscrit sísya and guru) were similar, with eventual right of succession. The Hindu sapinda, a family group of four generations, descendants of a common great-grandfather, seems to have the same significance and functions as the Irish derbfine and the Welsh gwely . The basic family unit was the same in both systems […] (Binchy, ‘The Linguistic and Historical Value of the Irish Law Tracts, 23, 27, 30, Proceedings of the British Academy, XXIX, 1943).
We can even claim agreement in the number of forms of marriage, form Binchy has suggested that two of the Irish ‘unions are a later development. (Studies in Early Irish Law, p.vi.)
Irish society was based on the family-group of four generations, and this may be assumed for the Celts. The wider unit was the tribe. In Caesars time there were about fifty tribes in Gaul, of which ony a few still recognised the institution of kingship. Most of them had adopted an oligarchical form of government. It is, therefore, from Irish sources, that we learn most about Celtic kingship. 
About the religion of the Celts much has been written, but the picture remains obscure. Caesar tells us that the Gauls worshipped Mercury, Apollo, Mars, Jupiter and Minerva. But he is simply equating Gaulish deities with his own Roman gods. He does not give us a single Gaulish name. Other historians do give us names, and many names occur in dedications all over the vast area once dominated by the Celts, from Galatia in Asia Minor to Spain and Britain. We have more than 400 names in all, and more than 300 of them occur only once. Perhaps these are names of local deities, each tribe or group of tribes having a special cult.
One important name is that of Lug, Welsh Llew. It occurs in dedications at Avranches in Switzerland and at Asma (Tarragona) in Spain as a plural, Lugoues, Lugouibus, and is common in the place-name Lugudunum, which is the name of Lyons, Loudon, Laon and Leon in France, of Leiden in Holland and Liegnitz in Silesia. Lugus must have been a great Celtic god, but we do not know what his special character was. In Ireland he was the god of Lugnasad, 1 August, which was evidently a harvest festival, and Lug may have been the god of fertility. At Lyons the feast of Augustus was celebrated on that day, apparently in substitution for the ancient feast of Lug. Máire MacNeill has shown that his feast was celebrated all over Ireland into our own times. (The Festival of Lughnasa, 1962; see also M. Tierney, ‘Lugh and Dionysius, in Éigse, x, p.265.) Lugs other name was Find, the Fairhaired One, and he survives on the Continent as Vindonnus, and in a few place-names: Uindobona (Vienna), Vindonissa, &c.
The poet Lucan says that Esus, Taranis and Teutates were the gods of the Celts, and two of these names are echoed in Welsh taran (Irish torann) thunder, and tud (Irish tuath) tribe. The faintness of the echo suggests a long interval in time between the first Celtic settlement of the British Isles and the date of our Gaulish evidence. Ogmios, the god of eloquence, described by Lucian, and Irish Ogma, who figures in the Battle of Moytura, are close in form, if not in function, although Thurneysen opposed the identification.
The cult of the Mothers was certainly Celtic, i.e. Mother Earth as a source of fertility, worshipped in triple form under the title matres or matronae . The dedications that survive on monuments in Gaul are in Latin, and the figures are sometimes represented with baskets of fruit or horns of plenty, or with children in their laps. There is a Welsh place-name rFoel Famau the hill of the mothers, which preserves the tradition. 
Another feature of Celtic religion that emerges clearly is that wells, rivers and sacred trees were objects of devotion and had patron gods or goddesses. Some rivers were themselves divine. There was a sanctuary of the Dea Sequana at the source of the river Seine, and one of the Dea Matrona near to the source of the Marne. In Ireland Bóinn (the Boyne) and Sinainn (the Shannon) were goddesses. Then there were animal gods: the bull, Taruos Trigaranus, who appears on the famous Paris monument in the Cluny museum [Plate 14]; the boar, Moccos; the goddess Epona, whose name suggests the cult of the horse; and the goddess Artio, perhaps a bear. The horned god Cernunnos appears on the Gundestrup Cauldron with a stag [Plate 4], and the goddess Damona suggests a cult of the cow.
The priests who maintained the cult of these gods were the druids, and Caesar tells us a good deal about them. They conducted the private and public sacrifices. They taught that the soul was immortal and passed after death into another body, rather as the Hindus believed. They also thought that all men were descended from Dis, the god of the Underworld, and they were learned in astronomy and natural philosophy).
After a victory, the captured animals were sacrificed to the god of war. Sometimes human victims were sacrificed, and Caesar says that criminals were preferred as being more pleasing to the gods, but that, failing them, others were chosen. This savage custom is mentioned by Lucan in a well-known passage where the gods Esus, Taranis and Teutates are said to be appeased by human sacrifice. (Lucan, Pharsalia, I, p.144.)
The Celts impressed the ancient historians as being impetuous and fearless in battle, easily roused and with a high sense of honour, arrogant in victory and desperate to the point of suicide in defeat, delighting in ornament, in feasting, in the recitation of poetry, admodum dedita religionibus much given to religion. Aristotle, in the Nichomachean Ethics says: We have no word for the man who is excessively fearless; perhaps one may call such a man mad or bereft of feeling, who fears nothing, neither earthquakes nor waves, as they say of the Celts (iii 7.7.). And Strabo quotes Ptolemy for the story that Celtic settlers on the Adriatic, when asked by Alexander the Great what they feared most, answered that it was lest the sky should fall. Cato the Elder says of the Celts of Cisalpine Gaul: Pleraque Gallia dual res industriosissime persequitur, rem militarem et argute loquithey have two great passions, to be brave in warfare and to speak well.
The picture that we get is of a people brave and gay, physically powerful, and amazingly successful in the early period. From Galatia in Asia Minor northwest to Scotland, and south again to Andalusia, one could travel in the third century BC without leaving Celtic territory. And although there was no empire, it was one culture.
Jacobsthal says that Celtic art, in all its variety and even though spread over so wide a territory, is one culture. And he adds:
Such were the ancestors of the peoples who emerge into history in the first centuries of the Christian era as Britanni and Hiberni, the Britons and the Irish.