of the poetical criticism is its tendency to be 'subjective,' that is, to reflect too strongly the personal prejudices of the author. It must virtually consist in giving the impression made upon the critic; and, however delicate his perception and wide his sympathy, he will be scarcely human if his judgments are not affected by his personal equation. No one could be more alive to the danger than Arnold, and his most characteristic teaching turns upon the mode of avoiding it. There are times, no doubt, when he relies too confidently upon the fineness of his perception, and then obviously has a slight spasm of diffidence. I have noticed how, in his Essays on Celtic Literature, he uses the true poetical or intuitive method: he recognises the precise point at which Shakespeare or Keats passes from the Greek to the Celtic note; he trusts to the fineness of his ear, like a musician who can detect the slightest discord. And we feel perhaps that a man who can decide, for example, an ethnological question by such means, who can by simple inspiration determine which are the Celtic and which are the Teutonic and which are the Norman elements in English character, is going a little beyond his tether. Arnold obviously feels so
