Now, as regards the character of the canons by which he criticises the works of other authors, in the majority of cases he employs simply his own geographical and military knowledge, showing, for instance, the impossibility in the accounts given of Nabis's march from Sparta simply by his acquaintance with the spots in question; or the inconsistency of those of the battle of Issus; or of the accounts given by Ephorus of the battles of Leuctra and Mantinea. In the latter case he says, if any one will take the trouble to measure out the ground of the site of the battle and then test the manoeuvres given, he will find how inaccurate the accounts are. In other cases he appeals to public documents, the importance of which he was always foremost in recognising; showing, for instance, by a document in the public archives of Rhodes how inaccurate were the accounts given of the battle of Lade by Zeno and Antisthenes. Or he appeals to psychological probability, rejecting, for instance, the scandalous stories told of Philip of Macedon, simply from the king's general greatness of character, and arguing that a boy so well educated and so respectably connected as Demochares (xii. 14) could never have been guilty of that of which evil rumour accused him. But the chief object of his literary censure is Timaeus, who had been so unsparing of his strictures on others. The general point which he makes against him, impugning his accuracy as a historian, is that he derived his knowledge of history not from the dangerous perils of a life of action but in the secure indolence of a narrow scholastic life. There is, indeed, no point on which he is so vehement as this. 'A history,' he says, 'written in a library gives as lifeless and as inaccurate a picture of history as a painting which is copied not from a living animal but from a stuffed one.' There is more difference, he says in another place, between the history of an eye-witness and that of one whose knowledge comes from books, than there is between the scenes of real life and the fictitious landscapes of theatrical scenery. Besides this, he enters into somewhat elaborate detailed criticism of passages where he thought Timaeus was following a wrong method and perverting truth, passages which it will be worth while to examine in detail. Timaeus, from the fact of there being a Roman custom to shoot a war-horse on a stated day, argued back to the Trojan origin of that people. Polybius, on the other hand, points out that the inference is quite unwarrantable, because horse-sacrifices are ordinary institutions common to all barbarous tribes. Timaeus here, as was so common with Greek writers, is arguing back from some custom of the present to an historical event in the past. Polybius really is employing the comparative method, showing how the custom was an ordinary step in the civilisation of every early people.