I have said that the Middle Ages inclined to the extreme; mediæval writers are, however, fond of insisting on “measure”; and this is almost inevitable in view of the large amount of classical, especially Aristotelian, survival throughout this period. But the two distinctively mediæval types, the saint and the knight, are neither of them mediators. They stand, however, on an entirely different footing as regards the law of measure. Not even Aristotle himself would maintain that the law of measure applies to saintliness, and in general to the religious realm. The saint in so far as he is saintly has undergone conversion, has in the literal sense of the word faced around and is looking in an entirely different direction from that to which the warnings “nothing too much” and “think as a mortal” apply. Very different psychic elements may indeed appear in any particular saint. A book has been published recently on the “Romanticism of St. Francis.” The truth seems to be that though St. Francis had his romantic side, he was even more religious than romantic. One may affirm with some confidence of another mediæval figure, Peter the Hermit, that he was, on the other hand, much more romantic than religious. For all the information we have tends to show that he was a very restless person and a man’s restlessness is ordinarily in inverse ratio to his religion. If the saint transcends in a way the law of measure, the knight on the other hand should be subject to it. For courage and the love of woman--his main interests in life--belong not to the religious but to the secular realm. But in his conception of love and courage the knight was plainly not a mediator but an extremist: he was haunted by the idea of adventure, of a love and courage that transcend the bounds not merely of the probable but of the possible. His imagination is romantic in the sense I have tried to define--it is straining, that is, beyond the confines of the real. Ruskin’s violent diatribe against Cervantes[147] for having killed “idealism” by his ridicule of these knightly exaggerations, is in itself absurd, but interesting as evidence of the quality of Ruskin’s own imagination. Like other romanticists I have cited, he seems to have been not unaware of his own kinship to Don Quixote. The very truth about either the mediæval or modern forms of romantic love--love which is on the secular level and at the same time sets itself above the law of measure--was uttered by Dr. Johnson in his comment on the heroic plays of Dryden: “By admitting the romantic omnipotence of love he has recommended as laudable and worthy of imitation that conduct which through all ages the good have censured as vicious and the bad have despised as foolish.”