In his preface to the _Cabinet of Antiques_, he explains how he dealt with his raw material. A young man has been prosecuted before the Assize Court, and had been condemned and branded. This case he connected with the story of an ancient family fallen from its high estate and dwelling in provincial surroundings. The story had dramatic elements in it, but less intensely dramatic than those of the young man's case. "This way of proceeding," he says, "should be that of an historian of manners and morals. His task consists in blending analogous facts in a single picture. Is he not rather bound to give the spirit than the letter of the happenings? He synthesizes them. Often it is necessary to pick out several similar characters in order to succeed in making up one, just as odd people are met with who are so ridiculous that two distinct persons may be created out of them. . . . . Literature uses a means employed in painting, which, to obtain a fine figure, adapts the hands of one model, the foot of another, the chest of a third, the shoulders of a fourth." The foregoing quotation raises the question of the significance of the term truth as applied to fiction. Evidently, it cannot have the same meaning as when applied to history or biography. In the latter, the writer invents neither circumstances nor actions, nor the persons engaged in them, but seeks to know the whole of the first two exactly as they occurred, and to interpret, as nearly to life as may be, the third. However, if he be a philosopher, he will perhaps try to show the intimate relations existing between these same persons and the events in which they were concerned; and, in doing so, he will step out of his proper role and assume one which is less easy for him than for the novelist to play, since the writer of fiction composes both his _dramatis personae_ and their story; and the concordance between them is more a matter of art than of science. Still it is possible that neither a novelist's characters nor their environment shall be in entire agreement with all observable facts. There may be arrangements, eliminations, additions, which, though pleasing to the reader, may remove the mimic world to a plane above that of the so-called real one. Thus removed, Balzac judged George Sand's production to be. And we must confess that, even in _Little Fadette_, _The Devil's Pool_, and _Francois le Champi_, it deals with human experience in a mode differing widely from that which the author of _Eugenie Grandet_ considered conform to truth.