In writing _Salammbo_ Flaubert set himself to renew the historical novel, as he had renewed the novel of manners. He would have admitted, doubtless, that perfect success in the historical novel is impossible, by the nature of the case. We are at best only half conscious of the reality of the things about us, only able to translate them approximately into any form of art. How much is left over, in the closest transcription of a mere line of houses in a street, of a passing steamer, of one's next-door neighbour, of the point of view of a foreigner looking along Piccadilly, of one's own state of mind, moment by moment, as one walks from Oxford Circus to the Marble Arch? Think, then, of the attempt to reconstruct no matter what period of the past, to distinguish the difference in the aspect of a world perhaps bossed with castles and ridged with ramparts, to two individualities encased within chain-armour! Flaubert chose his antiquity wisely: a period of which we know too little to confuse us, a city of which no stone is left on another, the minds of Barbarians who have left us no psychological documents. 'Be sure I have made no fantastic Carthage,' he says proudly, pointing to his documents; Ammianus Marcellinus, who has furnished him with 'the _exact_ form of a door'; the Bible and Theophrastus, from which he obtains his perfumes and his precious stones; Gresenius, from whom he gets his Punic names; the _Memoires de l'Academie des Inscriptions_. 'As for the temple of Tanit, I am sure of having reconstructed it as it was, with the treatise of the Syrian Goddess, with the medals of the Duc de Luynes, with what is known of the temple at Jerusalem, with a passage of St. Jerome, quoted by Seldon (_De Diis Syriis_), with the plan of the temple of Gozzo, which is quite Carthaginian, and best of all, with the ruins of the temple of Thugga, which I have seen myself, with my own eyes, and of which no traveller or antiquarian, so far as I know, has ever spoken.' But that, after all, as he admits (when, that is, he has proved point by point his minute accuracy to all that is known of ancient Carthage, his faithfulness to every indication which can serve for his guidance, his patience in grouping rather than his daring in the invention of action and details), that is not the question. 'I care little enough for archaeology! If the colour is not uniform, if the details are out of keeping, if the manners do not spring from the religion and the actions from the passions, if the characters are not consistent, if the costumes are not appropriate to the habits and the architecture to the climate, if, in a word, there is not harmony, I am in error. If not, no.'