The acute and learned author of _Les Maîtres de la Critique_ has, I think, been a little deceived by superficial characteristics of form and method. [Sidenote: _His critical quality and accomplishment._] A young man of twenty at the date of the _Cid_, and the battle over it, Saint-Evremond, fine gentleman as he was, no more forgot the forms of the quarrel than the attractions of the play. He has something a little scholastic, something of the earlier century, in his manner. Perhaps the best piece of criticism which he has left us—the _Dissertation sur le mot “vaste”_—recalls the known or alleged subjects of the earlier conversations of Malherbe and Racan, and their fellows, the later of Vaugelas and Chapelain and the young Academy. We have a formal _Jugement sur Senèque, Plutarque, et Pétrone_, a quite academic study of _Alexandre le Grand_, discussions of the character of tragedy generally, and the like. But the accomplished, agile, and independent spirit of the author is perpetually escaping from the restraints of his forms and models, and taking its own way according to its own taste. Perhaps, indeed, the fatal equivocation or ambiguity which seems to beset so many critical terms has worked here also: for to the present day the word “Taste” seems to excite quite dissimilar ideas in different minds. To some (as to Boileau and his followers it certainly did) it seems to suggest an antecedent law, a bar to which subjects are to be brought, something to which it is almost improper to apply the terms “good” or “bad,” because there is only one taste, and anything else is not taste at all but _un_taste. In this sense, though he might not have allowed it, I do not think that Saint-Evremond ever judged by “taste.” In the other—where taste means the approbation and satisfaction of a competent judge, well-gifted, well-tried, and taking pains to keep his palate clean—I think he always judged by it. That he often gave reasons for his judgment is nothing; one should _almost_ always do that. But one should always also remember that these reasons may be totally inapplicable to the next instance.