These Latin writers were contemporary witnesses, and among the captives taken by Roman armies they must have seen the men that they describe. Thus, in early times the Romans observed the same physical semblance in the two peoples, Celts and Germans. It may be pointed out, however, that the physical characteristics on which they lay stress are those which exhibit the greatest difference between these northern peoples and the peoples of southern Europe. For that reason we may suspect a certain element of exaggeration in the description. We may take leave to doubt whether all the Germans of antiquity were fair-haired and blue-eyed, as Tacitus describes them. It was the fair-haired and blue-eyed Germans and Celts that attracted the attention of Latin writers, accustomed to a population almost uniformly dark-haired and dark-eyed, and they would naturally seize upon the points of distinction and regard them as generally typical.

If, then, by the name Celts we cannot properly understand a distinct race, what are we to understand by it? By what criterion do we recognise any ancient population to have been Celts? The answer is undoubted—every ancient people that is known to have spoken any Celtic language is said to be a Celtic people. The term Celtic is indicative of language, not of race. We give the name Celts to the Irish and the Britons because we know that the ancient language of each people is a Celtic language.

A certain amount of enthusiasm, culminating in what is called Pan-Celticism, has gathered around the recognition of this fact that the Irish, the Gaels of Scotland, the Welsh and the Bretons are Celtic peoples. So much favour attached to the name Celtic that in our own time the Irish language was, so to speak, smuggled into the curricula of the Royal University and of the Intermediate Board under that name. What ancient writers called opus Hibernicum, "Irish work," is popularly known in Ireland as Celtic ornament. In the same way people speak of Celtic crosses, and there are even Celtic athletic clubs. There is no small amount of pride in the notion of being Celtic. It is somewhat remarkable, then, to find that throughout all their early history and tradition the Irish and the Britons alike show not the slightest atom of recognition that they were Celtic peoples. We do not find them acknowledging any kinship with the Gauls, or even with each other. In Christian times, their men of letters shaped out genealogical trees tracing the descent of each people from Japhet—and in these genealogies Gael and Briton and Gaul descend by lines as distinct as German and Greek. This absence of acknowledgment of kinship is all the more noteworthy because there is little reason to suppose that, before Latin displaced the Celtic speech of Gaul, the differences of dialect in the Celtic speech of Gaul, Britain, and Ireland were sufficient to prevent intercourse without interpreters.

From this ignorance of their Celtic kinship and origin we must draw one important conclusion. The extraordinary vitality of popular tradition in some respects must be set off by its extraordinary mortality in other respects. There must have been a time when the Celts of Ireland, Britain and Gaul were fully aware that they were nearer akin to each other than to the Germans and Italians, but this knowledge perished altogether from the popular memory and the popular consciousness.

It was re-discovered and re-established by a Scottish Gael, George Buchanan, in the sixteenth century. Buchanan, in his history of Scotland, published in 1589, dismissed as fabulous that section of the Irish and British genealogies that purported to trace the origin of each people, generation by generation, from Japhet. He was a man of great classical learning. No better refutation could be adduced of the notion that Bacon, who was a child when Buchanan wrote, established the inductive method of scientific proof than the clear and well-marshalled argument by which Buchanan proves from numerous Greek and Latin sources that the Gaels and the Britons were branches of the ancient Celtic people of the Continent.

An account of Buchanan's discourse on this subject will be found in an article by me in the "Irish Review," of December, 1913. Buchanan's discovery seems to have lain dormant, as regards any effect on learning or the popular mind, for more than a century. In his argument he dealt rather severely with the statements of a contemporary Welsh antiquary, Humphrey Llwyd, and this controversy had probably the effect of sowing the seed of what may be called Celtic consciousness in the soil of Welsh learning. In Ireland, though Buchanan's work was doubtless known and read, his theory of the Celtic origin of the Irish people and their language, and of their kinship to the Britons and the Continental Celts, does not appear to have been thought worth discussion, so firmly established were the ancient accounts which attributed to the Gaels of Ireland a Scythian origin. Yet these ancient accounts, as I propose to show in the third lecture of this series, did not belong to the true national tradition, ran counter to tradition, and owed their invention to the Latin learning of Ireland in the early Christian period.

In 1707 the publication of the first volume of Edward Llwyd's "Archæologia Britannica" exhibits the first fruiting of Buchanan's theory, in the form of a sort of conspectus of the Celtic languages then extant, namely, the Gaelic of Ireland and Scotland, and the British languages of Wales, Cornwall and Brittany. From this time onward, the existence of a group of Celtic peoples may be taken as a recognised fact in the learned world. I do not know whether anyone has yet traced the early stages of the recognition of the same fact in Continental learning.

The Celtic languages now began to attract attention from outside. I ought, however, to note here that already for a brief period the Irish language had seemed about to extend its influence beyond the limits of its own people. It will be remembered that Edmund Spenser, during his residence in Ireland (1586-1598), made some small acquaintance with Irish poetry which was translated for him, and that he was pleased in some degree with its peculiarities. About the same time an English official in Dublin reports to his masters in London that "the English in Dublin do now all speak Irish," and adds that they take a pleasure in speaking Irish. A primer of the Irish language was composed by the Baron of Delvin for the special use of Queen Elizabeth, and a facsimile of portion of it may be seen in Sir John Gilbert's "National Manuscripts of Ireland."

The growing interest in Celtic literature among outsiders is exemplified in some of the work of the English poet Gray, who died in 1771. His poem of "The Bard," reflected, if it did not initiate, the notions long afterwards fashionable of the character of the Celtic bards and of the spirit of their poetry. Gray had the reputation in his time of being an antiquarian. He made an English version of the vision-poem on the battle of Clontarf from the Icelandic saga of Burnt Njal, and from this same poem part of the inspiration of his "Bard" is acknowledged by him to have been derived. Gray also wrote English versions of some Welsh poems, and the novelty of poetic expression which he borrowed here seems to have baffled for once the critical experience of Johnson, who contents himself with saying that "the language is unlike the language of other poets." "The Bard" was published in 1755, and, if I am not mistaken, its weird rhapsodical spirit contained the germ of the Celtic literary revival, for Gray's "Bard" may be regarded as the literary parent of Macpherson's "Ossian." In 1760, five years after the publication of "The Bard," appeared the first collection of Macpherson's pretended translations, entitled "Fragments of Ancient Poetry Collected in the Highlands of Scotland." The consequences of this publication are fitly described by Dr. Magnus MacLean: "The arrival of James Macpherson marks a great moment in the history of Celtic literature. It was the signal for a general resurrection. It would seem as if he sounded the trumpet, and the graves of ancient manuscripts were opened, the books were read, and the dead were judged out of the things that were written in them." In 1764 was published Evans's "Specimens of the Poetry of the Ancient Welsh Bards"—which supplied Gray with fresh material. In 1784 appeared "Musical and Poetical Relics of the Welsh Bards," and from that time onward the stream of translations from Welsh to English was fairly continuous. Notwithstanding the controversy that soon arose about the authenticity of Macpherson's compositions, their direct influence and vogue went on increasing for half a century. Among those who shared in the Macpherson craze were Goethe and Napoleon Bonaparte. In France, de Villemarqué published his "Chants populaires de la Bretagne," a collection of poems from the Breton. In Scotland, Macpherson had several imitators. In Wales, the new movement took shape in the revival of the National Eisteddfod in 1819. In Ireland, the first fruits of Macpherson's genius are found in Walker's "Historical Memoirs of the Irish Bards," published in 1786, and in Charlotte Brooke's "Reliques of Irish Poetry," published in 1789. The originals in this case were genuine, including a number of poems of the kind called, since Macpherson's time, Ossianic.[1] The English versions supplied by Miss Brooke were in close imitation of the style and diction of Macpherson. The same influence extends to Hardiman's "Irish Minstrelsy," published in 1831.

[1] The Irish term for this class of poetry is "Fianaidheacht," and is of great antiquity.