In spite of this great fault, it must be allowed that Thucydides has surpassed all his rivals in the art of historical narration, in the art of producing an effect on the imagination, by skilful selection and disposition, without indulging in the license of invention. But narration, though an important part of the business of a historian, is not the whole. To append a moral to a work of fiction is either useless or superfluous. A fiction may give a more impressive effect to what is already known; but it can teach nothing new. If it presents to us characters and trains of events to which our experience furnishes us with nothing similar, instead of deriving instruction from it, we pronounce it unnatural. We do not form our opinions from it; but we try it by our preconceived opinions. Fiction, therefore, is essentially imitative. Its merit consists in its resemblance to a model with which we are already familiar, or to which at least we can instantly refer. Hence it is that the anecdotes which interest us most strongly in authentic narrative are offensive when introduced into novels; that what is called the romantic part of history is in fact the least romantic. It is delightful as history, because it contradicts our previous notions of human nature, and of the connection of causes and effects. It is, on that very account, shocking and incongruous in fiction. In fiction, the principles are given, to find the facts: in history, the facts are given, to find the principles; and the writer who does not explain the phenomena as well as state them, performs only one half of his office. Facts are the mere dross of history. It is from the abstract truth which interpenetrates them, and lies latent among them like gold in the ore, that the mass derives its whole value: and the precious particles are generally combined with the baser in such a manner that the separation is a task of the utmost difficulty. Here Thucydides is deficient: the deficiency, indeed, is not discreditable to him. It was the inevitable effect of circumstances. It was in the nature of things necessary that, in some part of its progress through political science, the human mind should reach that point which it attained in his time. Knowledge advances by steps, and not by leaps. The axioms of an English debating club would have been startling and mysterious paradoxes to the most enlightened statesmen of Athens. But it would be as absurd to speak contemptuously of the Athenian on this account as to ridicule Strabo for not having given us an account of Chili, or to talk of Ptolemy as we talk of Sir Richard Phillips. Still, when we wish for solid geographical information, we must prefer the solemn coxcombry of Pinkerton to the noble work of Strabo. If we wanted instruction respecting the solar system, we should consult the silliest girl from a boarding-school, rather than Ptolemy.