Ernest Renan described what he held to be Celtic characteristics in _The Poetry of the Celtic Races_. I must repeat the well-known sentences: 'No race communed so intimately as the Celtic race with the lower creation, or believed it to have so big a share of moral life.' The Celtic race had 'a realistic naturalism,' 'a love of nature for herself, a vivid feeling for her magic, commingled with the melancholy a man knows when he is face to face with her, and thinks he hears her communing with him about his origin and his destiny.' 'It has worn itself out in mistaking dreams for realities,' and 'compared with the classical imagination the Celtic imagination is indeed the infinite contrasted with the finite.' 'Its history is one long lament, it still recalls its exiles, its flights across the seas.' 'If at times it seems to be cheerful, its tear is not slow to glisten behind the smile. Its songs of joy end as elegies; there is nothing to equal the delightful sadness of its national melodies.' Matthew Arnold, in _The Study of Celtic Literature_, has accepted this passion for nature, this imaginativeness, this melancholy, as Celtic characteristics, but has described them more elaborately. The Celtic passion for nature comes almost more from a sense of her 'mystery' than of her 'beauty,' and it adds 'charm and magic' to nature, and the Celtic imaginativeness and melancholy are alike 'a passionate, turbulent, indomitable reaction against the despotism of fact.' The Celt is not melancholy, as Faust or Werther are melancholy, from 'a perfectly definite motive,' but because of something about him 'unaccountable, defiant and titanic.' How well one knows these sentences, better even than Renan's, and how well one knows the passages of prose and verse which he uses to prove that wherever English literature has the qualities these sentences describe, it has them from a Celtic source. Though I do not think any of us who write about Ireland have built any argument upon them, it is well to consider them a little, and see where they are helpful and where they are hurtful. If we do not, we may go mad some day, and the enemy root up our rose-garden and plant a cabbage-garden instead. Perhaps we must restate a little, Renan's and Arnold's argument. II