In this matter Livy allowed himself full liberty; and his work absorbed, and in a great measure blotted out, those of his predecessors. In his general preface he speaks of the two motives which animate new historians, as the hope that they will throw further light on events, or the belief that their own art will excel that of a ruder age. The former he hardly professes to do, at least as regards times anterior to his own; his hope is that by his pen the great story of the Republic will be told more impressively, more vividly, in a manner more stimulating to the reader and more worthy of the subject, than had hitherto been done. This purpose at least he amply and nobly carried out; nor can it be said to be a low ideal of the function of history. So far, however, as the office of the historian is to investigate facts, to get at the exact truth of what physically happened, or to appreciate the varying degrees of probability with which that truth can be attained, Livy falls far short of any respectable ideal. His romantic temper and the ethical bent of his mind alike indisposed him to set any very great value on facts as such. His history bears little trace of any independent investigation. Sources for history lay round him in immense profusion. The enormous collections made by Varro in every field of antiquarian research were at his hand, but he does not seem to have used them, still less to have undertaken any similar labour on his own account. While he never wilfully distorts the truth, he takes comparatively little pains to disengage it from fables and inaccuracies. In his account of a battle in Greece he finds that Valerius Antias puts the number of the enemy killed as inside ten thousand, while Claudius Quadrigarius says forty thousand. The discrepancy does not ruffle him, nor even seem to him very important; he contents himself with an expression of mild surprise that Valerius for once allows himself to be outstripped in exaggerating numbers. Yet where Valerius is his only authority or is not contradicted by others, he accepts his statements, figures and all, without uneasiness. This instance is typical of his method as a critical--or rather an uncritical --historian. When his authorities do not disagree, he accepts what they say without much question. When they do disagree, he has several courses open to him, and takes one or another according to his fancy at the moment. Sometimes he counts heads and follows the majority of his authors; sometimes he adopts the account of the earliest; often he tries to combine or mediate between discordant stories; when this is not easy, he chooses the account which is most superficially probable or most dramatically impressive. He even bases a choice on the ground that the story he adopts shows Roman statesmanship or virtue in a more favourable light, though he finds some of the inventions of Roman vanity too much for him to swallow. Throughout he tends to let his own preferences decide